diff --git a/README.md b/README.md
index d9f9bc7..10cfc36 100644
--- a/README.md
+++ b/README.md
@@ -19,8 +19,6 @@ This node package brings Meta's SAM3 model to ComfyUI, enabling:
 ![Video Segmentation Example](assets/video.png)
 *Example of semantic segmentation on video frames*
 
-> **Note**: Currently, this package supports **semantic segmentation** only.
-
 ## Features
 
 - 🖼️ **Image Segmentation**: Segment objects using natural language prompts
@@ -45,52 +43,74 @@ Load a SAM3 model for image or video segmentation.
 - `sam3_model`: Loaded SAM3 model for downstream nodes
 
 ### 2. SAM3 Image Segmentation
-Segment objects in images using text prompts.
+Segment objects in images using text prompts and optional geometric prompts.
 
 **Inputs:**
 - `sam3_model`: SAM3 model from Load SAM3 Model node
 - `images`: Input images to segment
 - `prompt`: Text description of objects to segment (e.g., "a cat", "person")
-- `threshold`: Confidence threshold for detections (0.0-1.0)
+- `threshold`: Confidence threshold for detections (0.0-1.0, default: 0.60)
 - `keep_model_loaded`: Keep model in VRAM after inference
 - `add_background`: Add background color (none, black, white, grey)
+- `coordinates_positive` (optional): Positive click coordinates to refine segmentation
+- `coordinates_negative` (optional): Negative click coordinates to exclude areas
+- `bboxes` (optional): Bounding boxes to guide segmentation
+- `mask` (optional): Input mask for refinement
 
 **Outputs:**
 - `masks`: Segmentation masks
 - `images`: Segmented images (with optional background)
+- `boxes`: Bounding box coordinates for detected objects
+- `scores`: Confidence scores for each detection
 
 ### 3. SAM3 Video Segmentation
-Track and segment objects across video frames.
+Track and segment objects across video frames with advanced prompting options.
 
 **Inputs:**
 - `sam3_model`: SAM3 model in video mode
-- `session_id`: Optional session ID to resume tracking
+- `session_id` (optional): Session ID to resume tracking from a previous session
 - `video_frames`: Video frames as image sequence
-- `prompt`: Text description of objects to track
-- `score_threshold_detection`: Detection confidence threshold
-- `new_det_thresh`: Threshold for adding new objects
+- `prompt`: Text description of objects to track (e.g., "person", "car")
+- `frame_index`: Frame where initial prompt is applied (0 to max frames)
+- `object_id`: Unique ID for multi-object tracking (1-1000, default: 1)
+- `score_threshold_detection`: Detection confidence threshold (0.0-1.0, default: 0.5)
+- `new_det_thresh`: Threshold for adding new objects (0.0-1.0, default: 0.7)
 - `propagation_direction`: Propagation direction (both, forward, backward)
-- `start_frame_index`: Frame index to start propagation
-- `keep_model_loaded`: Keep model in VRAM
-- `close_after_propagation`: Close session after completion
-- `extra_config`: Additional configuration from Extra Config node
+- `start_frame_index`: Frame index to start propagation (default: 0)
+- `max_frames_to_track`: Maximum frames to process (-1 for all frames)
+- `close_after_propagation`: Close session after completion (default: True)
+- `keep_model_loaded`: Keep model in VRAM after inference
+- `extra_config` (optional): Additional configuration from Extra Config node
+- `positive_coords` (optional): Positive click coordinates as JSON array
+- `negative_coords` (optional): Negative click coordinates as JSON array
+- `bbox` (optional): Bounding box to initialize tracking
 
 **Outputs:**
 - `masks`: Tracked segmentation masks for all frames
 - `session_id`: Session ID for resuming tracking
+- `objects`: Object tracking information and metadata
 
 ### 4. SAM3 Video Model Extra Config
-Configure advanced parameters for video segmentation.
+Configure advanced parameters for video segmentation to fine-tune tracking behavior.
 
-**Key Parameters:**
-- `assoc_iou_thresh`: IoU threshold for detection-to-track matching
-- `trk_assoc_iou_thresh`: Stricter IoU threshold for unmatched masklets
-- `hotstart_delay`: Delay outputs to remove unmatched/duplicate tracklets
-- `max_trk_keep_alive`: Maximum frames to keep track alive without detection
-- `det_nms_thresh`: IoU threshold for NMS
-- `fill_hole_area`: Fill holes in masks smaller than this area
-- `max_num_objects`: Maximum number of objects to track
-- And many more fine-tuning options...
+**Parameters:**
+- `assoc_iou_thresh`: IoU threshold for detection-to-track matching (0.0-1.0, default: 0.1)
+- `det_nms_thresh`: IoU threshold for detection NMS (0.0-1.0, default: 0.1)
+- `new_det_thresh`: Threshold for adding new objects (0.0-1.0, default: 0.7)
+- `hotstart_delay`: Hold off outputs for N frames to remove unmatched/duplicate tracklets (0-100, default: 15)
+- `hotstart_unmatch_thresh`: Remove tracklets unmatched for this many frames during hotstart (0-100, default: 8)
+- `hotstart_dup_thresh`: Remove overlapping tracklets during hotstart (0-100, default: 8)
+- `suppress_unmatched_within_hotstart`: Only suppress unmatched masks within hotstart period (default: True)
+- `min_trk_keep_alive`: Minimum keep-alive value (-100-0, default: -1, negative means immediate removal)
+- `max_trk_keep_alive`: Maximum frames to keep track alive without detections (0-100, default: 30)
+- `init_trk_keep_alive`: Initial keep-alive when new track is created (-10-100, default: 30)
+- `suppress_overlap_occlusion_thresh`: Threshold for suppressing overlapping objects (0.0-1.0, default: 0.7, 0.0 to disable)
+- `suppress_det_at_boundary`: Suppress detections close to image boundaries (default: False)
+- `fill_hole_area`: Fill holes in masks smaller than this area in pixels (0-1000, default: 16)
+- `recondition_every_nth_frame`: Recondition tracking every N frames (-1-1000, default: 16, -1 to disable)
+- `enable_masklet_confirmation`: Enable masklet confirmation to suppress unconfirmed tracklets (default: False)
+- `decrease_alive_for_empty_masks`: Decrease keep-alive counter for empty masklets (default: False)
+- `image_size`: Input image size for the model (256-2048, step: 8, default: 1008)
 
 **Output:**
 - `extra_config`: Configuration dictionary for Video Segmentation node
diff --git a/README_CN.md b/README_CN.md
index 0aee51d..792c4e1 100644
--- a/README_CN.md
+++ b/README_CN.md
@@ -19,7 +19,6 @@
 ![视频分割示例](assets/video.png)
 *视频帧语义分割示例*
 
-> **注意**：目前本插件仅支持了**语义分割**功能
 
 ## 功能特性
 
@@ -44,52 +43,74 @@
 - `sam3_model`：已加载的 SAM3 模型，供下游节点使用
 
 ### 2. SAM3 图像分割
-使用文本提示分割图像中的对象。
+使用文本提示和可选的几何提示分割图像中的对象。
 
 **输入：**
 - `sam3_model`：来自"加载 SAM3 模型"节点的 SAM3 模型
 - `images`：要分割的输入图像
 - `prompt`：要分割的对象的文本描述（例如："一只猫"、"人"）
-- `threshold`：检测的置信度阈值（0.0-1.0）
+- `threshold`：检测的置信度阈值（0.0-1.0，默认：0.60）
 - `keep_model_loaded`：推理后将模型保留在显存中
 - `add_background`：添加背景颜色（无、黑色、白色、灰色）
+- `coordinates_positive`（可选）：正向点击坐标以细化分割
+- `coordinates_negative`（可选）：负向点击坐标以排除区域
+- `bboxes`（可选）：边界框来引导分割
+- `mask`（可选）：用于细化的输入遮罩
 
 **输出：**
 - `masks`：分割遮罩
 - `images`：分割后的图像（可选背景）
+- `boxes`：检测到的对象的边界框坐标
+- `scores`：每个检测的置信度分数
 
 ### 3. SAM3 视频分割
-跨视频帧跟踪和分割对象。
+跨视频帧跟踪和分割对象，支持高级提示选项。
 
 **输入：**
 - `sam3_model`：视频模式的 SAM3 模型
-- `session_id`：可选的会话 ID，用于恢复跟踪
+- `session_id`（可选）：会话 ID，用于从之前的会话恢复跟踪
 - `video_frames`：作为图像序列的视频帧
-- `prompt`：要跟踪的对象的文本描述
-- `score_threshold_detection`：检测置信度阈值
-- `new_det_thresh`：添加新对象的阈值
-- `propagation_direction`：传播方向 (双向、前向、后向）
-- `start_frame_index`：开始传播的帧索引
-- `keep_model_loaded`：将模型保留在显存中
-- `close_after_propagation`：完成后关闭会话
-- `extra_config`：来自额外配置节点的附加配置
+- `prompt`：要跟踪的对象的文本描述（例如："人"、"汽车"）
+- `frame_index`：应用初始提示的帧位置（0 到最大帧数）
+- `object_id`：多对象跟踪的唯一 ID（1-1000，默认：1）
+- `score_threshold_detection`：检测置信度阈值（0.0-1.0，默认：0.5）
+- `new_det_thresh`：添加新对象的阈值（0.0-1.0，默认：0.7）
+- `propagation_direction`：传播方向（双向、前向、后向）
+- `start_frame_index`：开始传播的帧索引（默认：0）
+- `max_frames_to_track`：要处理的最大帧数（-1 表示所有帧）
+- `close_after_propagation`：完成后关闭会话（默认：True）
+- `keep_model_loaded`：推理后将模型保留在显存中
+- `extra_config`（可选）：来自额外配置节点的附加配置
+- `positive_coords`（可选）：正向点击坐标，JSON 数组格式
+- `negative_coords`（可选）：负向点击坐标，JSON 数组格式
+- `bbox`（可选）：用于初始化跟踪的边界框
 
 **输出：**
 - `masks`：所有帧的跟踪分割遮罩
 - `session_id`：用于恢复跟踪的会话 ID
+- `objects`：对象跟踪信息和元数据
 
 ### 4. SAM3 视频模型额外配置
-配置视频分割的高级参数。
+配置视频分割的高级参数，以微调跟踪行为。
 
-**主要参数：**
-- `assoc_iou_thresh`：检测到跟踪匹配的 IoU 阈值
-- `trk_assoc_iou_thresh`：不匹配掩码的更严格 IoU 阈值
-- `hotstart_delay`：延迟输出以移除不匹配/重复的轨迹
-- `max_trk_keep_alive`：没有检测时保持跟踪活动的最大帧数
-- `det_nms_thresh`：NMS 的 IoU 阈值
-- `fill_hole_area`：填充遮罩中小于此面积的孔洞
-- `max_num_objects`：要跟踪的最大对象数
-- 还有更多微调选项...
+**参数：**
+- `assoc_iou_thresh`：检测到跟踪匹配的 IoU 阈值（0.0-1.0，默认：0.1）
+- `det_nms_thresh`：检测 NMS 的 IoU 阈值（0.0-1.0，默认：0.1）
+- `new_det_thresh`：添加新对象的阈值（0.0-1.0，默认：0.7）
+- `hotstart_delay`：延迟 N 帧输出以移除不匹配/重复的轨迹（0-100，默认：15）
+- `hotstart_unmatch_thresh`：在热启动期间移除未匹配此帧数的轨迹（0-100，默认：8）
+- `hotstart_dup_thresh`：在热启动期间移除重叠的轨迹（0-100，默认：8）
+- `suppress_unmatched_within_hotstart`：仅在热启动期间抑制未匹配的遮罩（默认：True）
+- `min_trk_keep_alive`：最小保持活动值（-100-0，默认：-1，负值表示立即移除）
+- `max_trk_keep_alive`：没有检测时保持跟踪活动的最大帧数（0-100，默认：30）
+- `init_trk_keep_alive`：创建新跟踪时的初始保持活动值（-10-100，默认：30）
+- `suppress_overlap_occlusion_thresh`：基于近期遮挡抑制重叠对象的阈值（0.0-1.0，默认：0.7，0.0 表示禁用）
+- `suppress_det_at_boundary`：抑制接近图像边界的检测（默认：False）
+- `fill_hole_area`：填充遮罩中小于此面积（像素）的孔洞（0-1000，默认：16）
+- `recondition_every_nth_frame`：每 N 帧重新调整跟踪（-1-1000，默认：16，-1 表示禁用）
+- `enable_masklet_confirmation`：启用掩码确认以抑制未确认的轨迹（默认：False）
+- `decrease_alive_for_empty_masks`：减少空掩码的保持活动计数器（默认：False）
+- `image_size`：模型的输入图像大小（256-2048，步长：8，默认：1008）
 
 **输出：**
 - `extra_config`：视频分割节点的配置字典
diff --git a/assets/image.png b/assets/image.png
index 87c025f..489f204 100644
Binary files a/assets/image.png and b/assets/image.png differ
diff --git a/assets/video.png b/assets/video.png
index 36b83a9..fe5e0c1 100644
Binary files a/assets/video.png and b/assets/video.png differ
diff --git a/example_workflows/sam3_image_seg_by_points.json b/example_workflows/sam3_image_seg_by_points.json
new file mode 100644
index 0000000..4bc018e
--- /dev/null
+++ b/example_workflows/sam3_image_seg_by_points.json
@@ -0,0 +1,386 @@
+{
+  "id": "307743a2-8a4b-4090-b8dc-3125ff31fa14",
+  "revision": 0,
+  "last_node_id": 225,
+  "last_link_id": 311,
+  "nodes": [
+    {
+      "id": 222,
+      "type": "PreviewImage",
+      "pos": [
+        -3883.9588057841556,
+        593.0894010290336
+      ],
+      "size": [
+        339.2921196124935,
+        246
+      ],
+      "flags": {},
+      "order": 5,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "images",
+          "type": "IMAGE",
+          "link": 303
+        }
+      ],
+      "outputs": [],
+      "properties": {
+        "cnr_id": "comfy-core",
+        "ver": "0.3.70",
+        "Node name for S&R": "PreviewImage"
+      },
+      "widgets_values": []
+    },
+    {
+      "id": 218,
+      "type": "MaskPreview+",
+      "pos": [
+        -3880.8437110898844,
+        280.9015183807766
+      ],
+      "size": [
+        332.95740427927046,
+        258
+      ],
+      "flags": {},
+      "order": 4,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "mask",
+          "type": "MASK",
+          "link": 298
+        }
+      ],
+      "outputs": [],
+      "properties": {
+        "cnr_id": "comfyui_essentials",
+        "ver": "9d9f4bedfc9f0321c19faf71855e228c93bd0dc9",
+        "Node name for S&R": "MaskPreview+"
+      },
+      "widgets_values": []
+    },
+    {
+      "id": 190,
+      "type": "easy sam3ModelLoader",
+      "pos": [
+        -4245.32406706611,
+        379.2395155406056
+      ],
+      "size": [
+        314.09852906952574,
+        130
+      ],
+      "flags": {},
+      "order": 0,
+      "mode": 0,
+      "inputs": [],
+      "outputs": [
+        {
+          "name": "sam3_model",
+          "type": "EASY_SAM3_MODEL",
+          "links": [
+            282
+          ]
+        }
+      ],
+      "properties": {
+        "Node name for S&R": "easy sam3ModelLoader"
+      },
+      "widgets_values": [
+        "sam3.pt",
+        "image",
+        "cuda",
+        "fp16"
+      ]
+    },
+    {
+      "id": 213,
+      "type": "easy sam3ImageSegmentation",
+      "pos": [
+        -4251.231285188305,
+        584.8079182960263
+      ],
+      "size": [
+        337.2438563422779,
+        281.966544722883
+      ],
+      "flags": {},
+      "order": 3,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "sam3_model",
+          "type": "EASY_SAM3_MODEL",
+          "link": 282
+        },
+        {
+          "name": "images",
+          "type": "IMAGE",
+          "link": 311
+        },
+        {
+          "name": "coordinates_positive",
+          "shape": 7,
+          "type": "STRING",
+          "link": 307
+        },
+        {
+          "name": "coordinates_negative",
+          "shape": 7,
+          "type": "STRING",
+          "link": 309
+        },
+        {
+          "name": "bboxes",
+          "shape": 7,
+          "type": "BBOX",
+          "link": null
+        },
+        {
+          "name": "mask",
+          "shape": 7,
+          "type": "MASK",
+          "link": null
+        }
+      ],
+      "outputs": [
+        {
+          "name": "masks",
+          "shape": 6,
+          "type": "MASK",
+          "links": [
+            298
+          ]
+        },
+        {
+          "name": "images",
+          "shape": 6,
+          "type": "IMAGE",
+          "links": [
+            303
+          ]
+        },
+        {
+          "name": "boxes",
+          "shape": 6,
+          "type": "STRING",
+          "links": null
+        },
+        {
+          "name": "scores",
+          "shape": 6,
+          "type": "STRING",
+          "links": null
+        }
+      ],
+      "properties": {
+        "Node name for S&R": "easy sam3ImageSegmentation"
+      },
+      "widgets_values": [
+        "",
+        0.6,
+        false,
+        "none"
+      ]
+    },
+    {
+      "id": 224,
+      "type": "PointsEditor",
+      "pos": [
+        -4980.558137888023,
+        177.67803979107896
+      ],
+      "size": [
+        687,
+        780
+      ],
+      "flags": {},
+      "order": 2,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "bg_image",
+          "shape": 7,
+          "type": "IMAGE",
+          "link": 310
+        }
+      ],
+      "outputs": [
+        {
+          "name": "positive_coords",
+          "type": "STRING",
+          "links": [
+            307
+          ]
+        },
+        {
+          "name": "negative_coords",
+          "type": "STRING",
+          "links": [
+            309
+          ]
+        },
+        {
+          "name": "bbox",
+          "type": "BBOX",
+          "links": null
+        },
+        {
+          "name": "bbox_mask",
+          "type": "MASK",
+          "links": null
+        },
+        {
+          "name": "cropped_image",
+          "type": "IMAGE",
+          "links": null
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfyui-kjnodes",
+        "ver": "6ee278aa7d802c03ad713a7c5fbc02861b6773a8",
+        "Node name for S&R": "PointsEditor",
+        "points": "PointsEditor",
+        "neg_points": "PointsEditor",
+        "imgData": {
+          "name": "bg_image",
+          "base64": [
+            "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"
+          ]
+        }
+      },
+      "widgets_values": [
+        "{\"positive\":[{\"x\":323.6377618901521,\"y\":102.48529126521483}],\"negative\":[{\"x\":0,\"y\":0}]}",
+        "[{\"x\":323.6377618901521,\"y\":102.48529126521483}]",
+        "[{\"x\":0,\"y\":0}]",
+        "[{}]",
+        "[{}]",
+        "xyxy",
+        642,
+        480,
+        false,
+        "",
+        null
+      ]
+    },
+    {
+      "id": 221,
+      "type": "LoadImage",
+      "pos": [
+        -5225.148552188546,
+        518.7099633437581
+      ],
+      "size": [
+        221.39931371353123,
+        326
+      ],
+      "flags": {},
+      "order": 1,
+      "mode": 0,
+      "inputs": [],
+      "outputs": [
+        {
+          "name": "IMAGE",
+          "type": "IMAGE",
+          "links": [
+            310,
+            311
+          ]
+        },
+        {
+          "name": "MASK",
+          "type": "MASK",
+          "links": null
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfy-core",
+        "ver": "0.3.70",
+        "Node name for S&R": "LoadImage"
+      },
+      "widgets_values": [
+        "487faf25fc9bf7bde8a8583398aade7d.png",
+        "image"
+      ]
+    }
+  ],
+  "links": [
+    [
+      282,
+      190,
+      0,
+      213,
+      0,
+      "EASY_SAM3_MODEL"
+    ],
+    [
+      298,
+      213,
+      0,
+      218,
+      0,
+      "MASK"
+    ],
+    [
+      303,
+      213,
+      1,
+      222,
+      0,
+      "IMAGE"
+    ],
+    [
+      307,
+      224,
+      0,
+      213,
+      2,
+      "STRING"
+    ],
+    [
+      309,
+      224,
+      1,
+      213,
+      3,
+      "STRING"
+    ],
+    [
+      310,
+      221,
+      0,
+      224,
+      0,
+      "IMAGE"
+    ],
+    [
+      311,
+      221,
+      0,
+      213,
+      1,
+      "IMAGE"
+    ]
+  ],
+  "groups": [],
+  "config": {},
+  "extra": {
+    "ds": {
+      "scale": 0.9269622872433065,
+      "offset": [
+        5439.48041680524,
+        99.84517962436495
+      ]
+    },
+    "workflowRendererVersion": "LG",
+    "frontendVersion": "1.33.5",
+    "VHS_latentpreview": false,
+    "VHS_latentpreviewrate": 0,
+    "VHS_MetadataImage": true,
+    "VHS_KeepIntermediate": true
+  },
+  "version": 0.4
+}
\ No newline at end of file
diff --git a/example_workflows/sam3_image_seg_by_prompt.json b/example_workflows/sam3_image_seg_by_prompt.json
new file mode 100644
index 0000000..dcc1cbf
--- /dev/null
+++ b/example_workflows/sam3_image_seg_by_prompt.json
@@ -0,0 +1,281 @@
+{
+  "id": "307743a2-8a4b-4090-b8dc-3125ff31fa14",
+  "revision": 0,
+  "last_node_id": 223,
+  "last_link_id": 304,
+  "nodes": [
+    {
+      "id": 222,
+      "type": "PreviewImage",
+      "pos": [
+        -3883.9588057841556,
+        593.0894010290336
+      ],
+      "size": [
+        339.2921196124935,
+        246
+      ],
+      "flags": {},
+      "order": 4,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "images",
+          "type": "IMAGE",
+          "link": 303
+        }
+      ],
+      "outputs": [],
+      "properties": {
+        "cnr_id": "comfy-core",
+        "ver": "0.3.70",
+        "Node name for S&R": "PreviewImage"
+      },
+      "widgets_values": []
+    },
+    {
+      "id": 218,
+      "type": "MaskPreview+",
+      "pos": [
+        -3880.8437110898844,
+        280.9015183807766
+      ],
+      "size": [
+        332.95740427927046,
+        258
+      ],
+      "flags": {},
+      "order": 3,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "mask",
+          "type": "MASK",
+          "link": 298
+        }
+      ],
+      "outputs": [],
+      "properties": {
+        "cnr_id": "comfyui_essentials",
+        "ver": "9d9f4bedfc9f0321c19faf71855e228c93bd0dc9",
+        "Node name for S&R": "MaskPreview+"
+      },
+      "widgets_values": []
+    },
+    {
+      "id": 190,
+      "type": "easy sam3ModelLoader",
+      "pos": [
+        -4235.041825672723,
+        334.7519653807837
+      ],
+      "size": [
+        314.09852906952574,
+        130
+      ],
+      "flags": {},
+      "order": 0,
+      "mode": 0,
+      "inputs": [],
+      "outputs": [
+        {
+          "name": "sam3_model",
+          "type": "EASY_SAM3_MODEL",
+          "links": [
+            282
+          ]
+        }
+      ],
+      "properties": {
+        "Node name for S&R": "easy sam3ModelLoader"
+      },
+      "widgets_values": [
+        "sam3.pt",
+        "image",
+        "cuda",
+        "fp16"
+      ]
+    },
+    {
+      "id": 221,
+      "type": "LoadImage",
+      "pos": [
+        -4678.319118164902,
+        372.4935004943305
+      ],
+      "size": [
+        406.60579628243977,
+        446.341753250938
+      ],
+      "flags": {},
+      "order": 1,
+      "mode": 0,
+      "inputs": [],
+      "outputs": [
+        {
+          "name": "IMAGE",
+          "type": "IMAGE",
+          "links": [
+            302
+          ]
+        },
+        {
+          "name": "MASK",
+          "type": "MASK",
+          "links": null
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfy-core",
+        "ver": "0.3.70",
+        "Node name for S&R": "LoadImage"
+      },
+      "widgets_values": [
+        "1 (3).png",
+        "image"
+      ]
+    },
+    {
+      "id": 213,
+      "type": "easy sam3ImageSegmentation",
+      "pos": [
+        -4241.387303264146,
+        538.9760914950197
+      ],
+      "size": [
+        337.2438563422779,
+        281.966544722883
+      ],
+      "flags": {},
+      "order": 2,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "sam3_model",
+          "type": "EASY_SAM3_MODEL",
+          "link": 282
+        },
+        {
+          "name": "images",
+          "type": "IMAGE",
+          "link": 302
+        },
+        {
+          "name": "coordinates_positive",
+          "shape": 7,
+          "type": "STRING",
+          "link": null
+        },
+        {
+          "name": "coordinates_negative",
+          "shape": 7,
+          "type": "STRING",
+          "link": null
+        },
+        {
+          "name": "bboxes",
+          "shape": 7,
+          "type": "BBOX",
+          "link": null
+        },
+        {
+          "name": "mask",
+          "shape": 7,
+          "type": "MASK",
+          "link": null
+        }
+      ],
+      "outputs": [
+        {
+          "name": "masks",
+          "shape": 6,
+          "type": "MASK",
+          "links": [
+            298
+          ]
+        },
+        {
+          "name": "images",
+          "shape": 6,
+          "type": "IMAGE",
+          "links": [
+            303
+          ]
+        },
+        {
+          "name": "boxes",
+          "shape": 6,
+          "type": "STRING",
+          "links": null
+        },
+        {
+          "name": "scores",
+          "shape": 6,
+          "type": "STRING",
+          "links": null
+        }
+      ],
+      "properties": {
+        "Node name for S&R": "easy sam3ImageSegmentation"
+      },
+      "widgets_values": [
+        "apple",
+        0.4,
+        false,
+        "none"
+      ]
+    }
+  ],
+  "links": [
+    [
+      282,
+      190,
+      0,
+      213,
+      0,
+      "EASY_SAM3_MODEL"
+    ],
+    [
+      298,
+      213,
+      0,
+      218,
+      0,
+      "MASK"
+    ],
+    [
+      302,
+      221,
+      0,
+      213,
+      1,
+      "IMAGE"
+    ],
+    [
+      303,
+      213,
+      1,
+      222,
+      0,
+      "IMAGE"
+    ]
+  ],
+  "groups": [],
+  "config": {},
+  "extra": {
+    "ds": {
+      "scale": 1.3920191078517683,
+      "offset": [
+        4898.859436279234,
+        -79.3314399803767
+      ]
+    },
+    "workflowRendererVersion": "LG",
+    "frontendVersion": "1.33.5",
+    "VHS_latentpreview": false,
+    "VHS_latentpreviewrate": 0,
+    "VHS_MetadataImage": true,
+    "VHS_KeepIntermediate": true
+  },
+  "version": 0.4
+}
\ No newline at end of file
diff --git a/example_workflows/sam3_video_seg_by_point.json b/example_workflows/sam3_video_seg_by_point.json
new file mode 100644
index 0000000..3727945
--- /dev/null
+++ b/example_workflows/sam3_video_seg_by_point.json
@@ -0,0 +1,487 @@
+{
+  "id": "307743a2-8a4b-4090-b8dc-3125ff31fa14",
+  "revision": 0,
+  "last_node_id": 232,
+  "last_link_id": 319,
+  "nodes": [
+    {
+      "id": 226,
+      "type": "VHS_LoadVideo",
+      "pos": [
+        -5383.15775926406,
+        159.3386979191607
+      ],
+      "size": [
+        237.6240234375,
+        701.2149739583333
+      ],
+      "flags": {},
+      "order": 0,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "meta_batch",
+          "shape": 7,
+          "type": "VHS_BatchManager",
+          "link": null
+        },
+        {
+          "name": "vae",
+          "shape": 7,
+          "type": "VAE",
+          "link": null
+        }
+      ],
+      "outputs": [
+        {
+          "name": "IMAGE",
+          "type": "IMAGE",
+          "links": [
+            313,
+            314
+          ]
+        },
+        {
+          "name": "frame_count",
+          "type": "INT",
+          "links": null
+        },
+        {
+          "name": "audio",
+          "type": "AUDIO",
+          "links": null
+        },
+        {
+          "name": "video_info",
+          "type": "VHS_VIDEOINFO",
+          "links": null
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfyui-videohelpersuite",
+        "ver": "8923bd836bdab8b7bbdf4ed104b7d045e70c66e2",
+        "Node name for S&R": "VHS_LoadVideo"
+      },
+      "widgets_values": {
+        "video": "时尚走秀.mp4",
+        "force_rate": 24,
+        "custom_width": 480,
+        "custom_height": 832,
+        "frame_load_cap": 121,
+        "skip_first_frames": 0,
+        "select_every_nth": 1,
+        "format": "AnimateDiff",
+        "choose video to upload": "image",
+        "videopreview": {
+          "hidden": false,
+          "paused": false,
+          "params": {
+            "filename": "时尚走秀.mp4",
+            "type": "input",
+            "format": "video/mp4",
+            "force_rate": 24,
+            "custom_width": 480,
+            "custom_height": 832,
+            "frame_load_cap": 121,
+            "skip_first_frames": 0,
+            "select_every_nth": 1
+          }
+        }
+      }
+    },
+    {
+      "id": 228,
+      "type": "easy sam3ModelLoader",
+      "pos": [
+        -4519.153385861623,
+        232.95641658512474
+      ],
+      "size": [
+        270,
+        130
+      ],
+      "flags": {},
+      "order": 1,
+      "mode": 0,
+      "inputs": [],
+      "outputs": [
+        {
+          "name": "sam3_model",
+          "type": "EASY_SAM3_MODEL",
+          "links": [
+            312
+          ]
+        }
+      ],
+      "properties": {
+        "Node name for S&R": "easy sam3ModelLoader"
+      },
+      "widgets_values": [
+        "sam3.pt",
+        "video",
+        "cuda",
+        "fp16"
+      ]
+    },
+    {
+      "id": 227,
+      "type": "easy sam3VideoSegmentation",
+      "pos": [
+        -4529.19864967506,
+        443.5309937256167
+      ],
+      "size": [
+        314.00554584873043,
+        437.1614502609848
+      ],
+      "flags": {},
+      "order": 3,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "sam3_model",
+          "type": "EASY_SAM3_MODEL",
+          "link": 312
+        },
+        {
+          "name": "video_frames",
+          "type": "IMAGE",
+          "link": 313
+        },
+        {
+          "name": "session_id",
+          "shape": 7,
+          "type": "STRING",
+          "link": null
+        },
+        {
+          "name": "extra_config",
+          "shape": 7,
+          "type": "EASY_SAM3_EXTRA_CONFIG",
+          "link": null
+        },
+        {
+          "name": "positive_coords",
+          "shape": 7,
+          "type": "STRING",
+          "link": 316
+        },
+        {
+          "name": "negative_coords",
+          "shape": 7,
+          "type": "STRING",
+          "link": 317
+        },
+        {
+          "name": "bbox",
+          "shape": 7,
+          "type": "BBOX",
+          "link": null
+        }
+      ],
+      "outputs": [
+        {
+          "name": "masks",
+          "type": "MASK",
+          "links": [
+            318
+          ]
+        },
+        {
+          "name": "session_id",
+          "type": "STRING",
+          "links": null
+        },
+        {
+          "name": "objects",
+          "type": "EASY_SAM3_OBJECTS_OUTPUT",
+          "links": null
+        }
+      ],
+      "properties": {
+        "Node name for S&R": "easy sam3VideoSegmentation"
+      },
+      "widgets_values": [
+        "",
+        0,
+        1,
+        0.5,
+        0.7,
+        "both",
+        0,
+        -1,
+        true,
+        false
+      ]
+    },
+    {
+      "id": 231,
+      "type": "MaskToImage",
+      "pos": [
+        -4446.76917913515,
+        119.15114245007179
+      ],
+      "size": [
+        184.6236328125,
+        26
+      ],
+      "flags": {},
+      "order": 4,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "mask",
+          "type": "MASK",
+          "link": 318
+        }
+      ],
+      "outputs": [
+        {
+          "name": "IMAGE",
+          "type": "IMAGE",
+          "links": [
+            319
+          ]
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfy-core",
+        "ver": "0.3.70",
+        "Node name for S&R": "MaskToImage"
+      }
+    },
+    {
+      "id": 229,
+      "type": "PointsEditor",
+      "pos": [
+        -5086.056934977531,
+        -161.88694241011603
+      ],
+      "size": [
+        525,
+        1132
+      ],
+      "flags": {},
+      "order": 2,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "bg_image",
+          "shape": 7,
+          "type": "IMAGE",
+          "link": 314
+        }
+      ],
+      "outputs": [
+        {
+          "name": "positive_coords",
+          "type": "STRING",
+          "links": [
+            316
+          ]
+        },
+        {
+          "name": "negative_coords",
+          "type": "STRING",
+          "links": [
+            317
+          ]
+        },
+        {
+          "name": "bbox",
+          "type": "BBOX",
+          "links": null
+        },
+        {
+          "name": "bbox_mask",
+          "type": "MASK",
+          "links": null
+        },
+        {
+          "name": "cropped_image",
+          "type": "IMAGE",
+          "links": []
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfyui-kjnodes",
+        "ver": "6ee278aa7d802c03ad713a7c5fbc02861b6773a8",
+        "Node name for S&R": "PointsEditor",
+        "points": "PointsEditor",
+        "neg_points": "PointsEditor",
+        "imgData": {
+          "name": "bg_image",
+          "base64": [
+            "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"
+          ]
+        }
+      },
+      "widgets_values": [
+        "{\"positive\":[{\"x\":234.54790743932173,\"y\":266.5317129992292},{\"x\":176.00976209894242,\"y\":433.6134536477257}],\"negative\":[{\"x\":0,\"y\":0}]}",
+        "[{\"x\":234.54790743932173,\"y\":266.5317129992292},{\"x\":176.00976209894242,\"y\":433.6134536477257}]",
+        "[{\"x\":0,\"y\":0}]",
+        "[{}]",
+        "[{}]",
+        "xyxy",
+        480,
+        832,
+        false,
+        "",
+        null
+      ]
+    },
+    {
+      "id": 232,
+      "type": "VHS_VideoCombine",
+      "pos": [
+        -4186.105715232003,
+        -114.75297305036214
+      ],
+      "size": [
+        422.65873186131785,
+        1045.941801892951
+      ],
+      "flags": {},
+      "order": 5,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "images",
+          "type": "IMAGE",
+          "link": 319
+        },
+        {
+          "name": "audio",
+          "shape": 7,
+          "type": "AUDIO",
+          "link": null
+        },
+        {
+          "name": "meta_batch",
+          "shape": 7,
+          "type": "VHS_BatchManager",
+          "link": null
+        },
+        {
+          "name": "vae",
+          "shape": 7,
+          "type": "VAE",
+          "link": null
+        }
+      ],
+      "outputs": [
+        {
+          "name": "Filenames",
+          "type": "VHS_FILENAMES",
+          "links": null
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfyui-videohelpersuite",
+        "ver": "8923bd836bdab8b7bbdf4ed104b7d045e70c66e2",
+        "Node name for S&R": "VHS_VideoCombine"
+      },
+      "widgets_values": {
+        "frame_rate": 24,
+        "loop_count": 0,
+        "filename_prefix": "AnimateDiff",
+        "format": "video/h264-mp4",
+        "pix_fmt": "yuv420p",
+        "crf": 19,
+        "save_metadata": false,
+        "trim_to_audio": false,
+        "pingpong": false,
+        "save_output": false,
+        "videopreview": {
+          "hidden": false,
+          "paused": false,
+          "params": {
+            "filename": "AnimateDiff_00002.mp4",
+            "subfolder": "",
+            "type": "temp",
+            "format": "video/h264-mp4",
+            "frame_rate": 24,
+            "workflow": "AnimateDiff_00002.png",
+            "fullpath": "E:\\ComfyUI_windows_portable\\ComfyUI\\temp\\AnimateDiff_00002.mp4"
+          }
+        }
+      }
+    }
+  ],
+  "links": [
+    [
+      312,
+      228,
+      0,
+      227,
+      0,
+      "EASY_SAM3_MODEL"
+    ],
+    [
+      313,
+      226,
+      0,
+      227,
+      1,
+      "IMAGE"
+    ],
+    [
+      314,
+      226,
+      0,
+      229,
+      0,
+      "IMAGE"
+    ],
+    [
+      316,
+      229,
+      0,
+      227,
+      4,
+      "STRING"
+    ],
+    [
+      317,
+      229,
+      1,
+      227,
+      5,
+      "STRING"
+    ],
+    [
+      318,
+      227,
+      0,
+      231,
+      0,
+      "MASK"
+    ],
+    [
+      319,
+      231,
+      0,
+      232,
+      0,
+      "IMAGE"
+    ]
+  ],
+  "groups": [],
+  "config": {},
+  "extra": {
+    "ds": {
+      "scale": 0.8579069603827805,
+      "offset": [
+        5761.6061773065,
+        340.88016715010775
+      ]
+    },
+    "workflowRendererVersion": "LG",
+    "frontendVersion": "1.33.5",
+    "VHS_latentpreview": false,
+    "VHS_latentpreviewrate": 0,
+    "VHS_MetadataImage": true,
+    "VHS_KeepIntermediate": true
+  },
+  "version": 0.4
+}
\ No newline at end of file
diff --git a/example_workflows/sam3_video_seg_by_prompt.json b/example_workflows/sam3_video_seg_by_prompt.json
new file mode 100644
index 0000000..6c32a96
--- /dev/null
+++ b/example_workflows/sam3_video_seg_by_prompt.json
@@ -0,0 +1,382 @@
+{
+  "id": "307743a2-8a4b-4090-b8dc-3125ff31fa14",
+  "revision": 0,
+  "last_node_id": 232,
+  "last_link_id": 319,
+  "nodes": [
+    {
+      "id": 228,
+      "type": "easy sam3ModelLoader",
+      "pos": [
+        -4519.153385861623,
+        232.95641658512474
+      ],
+      "size": [
+        270,
+        130
+      ],
+      "flags": {},
+      "order": 0,
+      "mode": 0,
+      "inputs": [],
+      "outputs": [
+        {
+          "name": "sam3_model",
+          "type": "EASY_SAM3_MODEL",
+          "links": [
+            312
+          ]
+        }
+      ],
+      "properties": {
+        "Node name for S&R": "easy sam3ModelLoader"
+      },
+      "widgets_values": [
+        "sam3.pt",
+        "video",
+        "cuda",
+        "fp16"
+      ]
+    },
+    {
+      "id": 231,
+      "type": "MaskToImage",
+      "pos": [
+        -4446.76917913515,
+        119.15114245007179
+      ],
+      "size": [
+        184.6236328125,
+        26
+      ],
+      "flags": {},
+      "order": 3,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "mask",
+          "type": "MASK",
+          "link": 318
+        }
+      ],
+      "outputs": [
+        {
+          "name": "IMAGE",
+          "type": "IMAGE",
+          "links": [
+            319
+          ]
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfy-core",
+        "ver": "0.3.70",
+        "Node name for S&R": "MaskToImage"
+      }
+    },
+    {
+      "id": 232,
+      "type": "VHS_VideoCombine",
+      "pos": [
+        -4186.105715232003,
+        -114.75297305036214
+      ],
+      "size": [
+        422.65873186131785,
+        1045.941801892951
+      ],
+      "flags": {},
+      "order": 4,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "images",
+          "type": "IMAGE",
+          "link": 319
+        },
+        {
+          "name": "audio",
+          "shape": 7,
+          "type": "AUDIO",
+          "link": null
+        },
+        {
+          "name": "meta_batch",
+          "shape": 7,
+          "type": "VHS_BatchManager",
+          "link": null
+        },
+        {
+          "name": "vae",
+          "shape": 7,
+          "type": "VAE",
+          "link": null
+        }
+      ],
+      "outputs": [
+        {
+          "name": "Filenames",
+          "type": "VHS_FILENAMES",
+          "links": null
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfyui-videohelpersuite",
+        "ver": "8923bd836bdab8b7bbdf4ed104b7d045e70c66e2",
+        "Node name for S&R": "VHS_VideoCombine"
+      },
+      "widgets_values": {
+        "frame_rate": 24,
+        "loop_count": 0,
+        "filename_prefix": "AnimateDiff",
+        "format": "video/h264-mp4",
+        "pix_fmt": "yuv420p",
+        "crf": 19,
+        "save_metadata": false,
+        "trim_to_audio": false,
+        "pingpong": false,
+        "save_output": false,
+        "videopreview": {
+          "hidden": false,
+          "paused": false,
+          "params": {
+            "filename": "AnimateDiff_00003.mp4",
+            "subfolder": "",
+            "type": "temp",
+            "format": "video/h264-mp4",
+            "frame_rate": 24,
+            "workflow": "AnimateDiff_00003.png",
+            "fullpath": "E:\\ComfyUI_windows_portable\\ComfyUI\\temp\\AnimateDiff_00003.mp4"
+          }
+        }
+      }
+    },
+    {
+      "id": 226,
+      "type": "VHS_LoadVideo",
+      "pos": [
+        -4808.758351453994,
+        184.48164872396106
+      ],
+      "size": [
+        237.6240234375,
+        701.2149739583333
+      ],
+      "flags": {},
+      "order": 1,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "meta_batch",
+          "shape": 7,
+          "type": "VHS_BatchManager",
+          "link": null
+        },
+        {
+          "name": "vae",
+          "shape": 7,
+          "type": "VAE",
+          "link": null
+        }
+      ],
+      "outputs": [
+        {
+          "name": "IMAGE",
+          "type": "IMAGE",
+          "links": [
+            313
+          ]
+        },
+        {
+          "name": "frame_count",
+          "type": "INT",
+          "links": null
+        },
+        {
+          "name": "audio",
+          "type": "AUDIO",
+          "links": null
+        },
+        {
+          "name": "video_info",
+          "type": "VHS_VIDEOINFO",
+          "links": null
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfyui-videohelpersuite",
+        "ver": "8923bd836bdab8b7bbdf4ed104b7d045e70c66e2",
+        "Node name for S&R": "VHS_LoadVideo"
+      },
+      "widgets_values": {
+        "video": "时尚走秀.mp4",
+        "force_rate": 24,
+        "custom_width": 480,
+        "custom_height": 832,
+        "frame_load_cap": 121,
+        "skip_first_frames": 0,
+        "select_every_nth": 1,
+        "format": "AnimateDiff",
+        "choose video to upload": "image",
+        "videopreview": {
+          "hidden": false,
+          "paused": false,
+          "params": {
+            "filename": "时尚走秀.mp4",
+            "type": "input",
+            "format": "video/mp4",
+            "force_rate": 24,
+            "custom_width": 480,
+            "custom_height": 832,
+            "frame_load_cap": 121,
+            "skip_first_frames": 0,
+            "select_every_nth": 1
+          }
+        }
+      }
+    },
+    {
+      "id": 227,
+      "type": "easy sam3VideoSegmentation",
+      "pos": [
+        -4529.19864967506,
+        443.5309937256167
+      ],
+      "size": [
+        314.00554584873043,
+        437.1614502609848
+      ],
+      "flags": {},
+      "order": 2,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "sam3_model",
+          "type": "EASY_SAM3_MODEL",
+          "link": 312
+        },
+        {
+          "name": "video_frames",
+          "type": "IMAGE",
+          "link": 313
+        },
+        {
+          "name": "session_id",
+          "shape": 7,
+          "type": "STRING",
+          "link": null
+        },
+        {
+          "name": "extra_config",
+          "shape": 7,
+          "type": "EASY_SAM3_EXTRA_CONFIG",
+          "link": null
+        },
+        {
+          "name": "positive_coords",
+          "shape": 7,
+          "type": "STRING",
+          "link": null
+        },
+        {
+          "name": "negative_coords",
+          "shape": 7,
+          "type": "STRING",
+          "link": null
+        },
+        {
+          "name": "bbox",
+          "shape": 7,
+          "type": "BBOX",
+          "link": null
+        }
+      ],
+      "outputs": [
+        {
+          "name": "masks",
+          "type": "MASK",
+          "links": [
+            318
+          ]
+        },
+        {
+          "name": "session_id",
+          "type": "STRING",
+          "links": null
+        },
+        {
+          "name": "objects",
+          "type": "EASY_SAM3_OBJECTS_OUTPUT",
+          "links": null
+        }
+      ],
+      "properties": {
+        "Node name for S&R": "easy sam3VideoSegmentation"
+      },
+      "widgets_values": [
+        "dress",
+        0,
+        1,
+        0.5,
+        0.7,
+        "both",
+        0,
+        -1,
+        true,
+        false
+      ]
+    }
+  ],
+  "links": [
+    [
+      312,
+      228,
+      0,
+      227,
+      0,
+      "EASY_SAM3_MODEL"
+    ],
+    [
+      313,
+      226,
+      0,
+      227,
+      1,
+      "IMAGE"
+    ],
+    [
+      318,
+      227,
+      0,
+      231,
+      0,
+      "MASK"
+    ],
+    [
+      319,
+      231,
+      0,
+      232,
+      0,
+      "IMAGE"
+    ]
+  ],
+  "groups": [],
+  "config": {},
+  "extra": {
+    "ds": {
+      "scale": 0.8579069603827805,
+      "offset": [
+        5332.017781339226,
+        351.4800928063147
+      ]
+    },
+    "workflowRendererVersion": "LG",
+    "frontendVersion": "1.33.5",
+    "VHS_latentpreview": false,
+    "VHS_latentpreviewrate": 0,
+    "VHS_MetadataImage": true,
+    "VHS_KeepIntermediate": true
+  },
+  "version": 0.4
+}
\ No newline at end of file
diff --git a/locales/en/nodeDefs.json b/locales/en/nodeDefs.json
index 276d5ad..d867eab 100644
--- a/locales/en/nodeDefs.json
+++ b/locales/en/nodeDefs.json
@@ -51,6 +51,22 @@
       "add_background": {
         "name": "add_background",
         "tooltip": "Add background color to segmented images"
+      },
+      "coordinates_positive": {
+        "name": "coordinates_positive",
+        "tooltip": "Positive click coordinates for refinement"
+      },
+      "coordinates_negative": {
+        "name": "coordinates_negative",
+        "tooltip": "Negative click coordinates for refinement"
+      },
+      "bboxes": {
+        "name": "bboxes",
+        "tooltip": "Bounding boxes for object detection"
+      },
+      "mask": {
+        "name": "mask",
+        "tooltip": "Input mask for refinement"
       }
     },
     "outputs": {
@@ -61,6 +77,14 @@
       "1": {
         "name": "images",
         "tooltip": "Segmentation images"
+      },
+      "2": {
+        "name": "boxes",
+        "tooltip": "Detected bounding boxes"
+      },
+      "3": {
+        "name": "scores",
+        "tooltip": "Detection confidence scores"
       }
     }
   },
@@ -83,6 +107,14 @@
         "name": "prompt",
         "tooltip": "Text description of objects to track (e.g., 'person', 'car')"
       },
+      "frame_index": {
+        "name": "frame_index",
+        "tooltip": "Frame where initial prompt is applied"
+      },
+      "object_id": {
+        "name": "object_id",
+        "tooltip": "Unique ID for multi-object tracking"
+      },
       "score_threshold_detection": {
         "name": "score_threshold_detection",
         "tooltip": "Confidence threshold for detections, default is 0.5"
@@ -91,33 +123,41 @@
         "name": "new_det_thresh",
         "tooltip": "Threshold for a detection to be added as a new object, default is 0.7"
       },
-      "object_frame_index": {
-        "name": "object_frame_index",
-        "tooltip": "Object Frame index to start tracking from"
-      },
-      "object_id": {
-        "name": "object_id",
-        "tooltip": "Object ID to track (1-100)"
-      },
-      "start_to_propagate": {
-        "name": "start_to_propagate",
-        "tooltip": "Propagation direction: disabled, both, forward, or backward"
+      "propagation_direction": {
+        "name": "propagation_direction",
+        "tooltip": "Propagation direction: both, forward, or backward"
       },
       "start_frame_index": {
         "name": "start_frame_index",
         "tooltip": "Frame index to start propagation from"
       },
-      "keep_model_loaded": {
-        "name": "keep_model_loaded",
-        "tooltip": "Keep model in VRAM after inference"
+      "max_frames_to_track": {
+        "name": "max_frames_to_track",
+        "tooltip": "Advanced: Max frames to process (-1 for all)"
       },
       "close_after_propagation": {
         "name": "close_after_propagation",
         "tooltip": "Close the session after propagation"
       },
+      "keep_model_loaded": {
+        "name": "keep_model_loaded",
+        "tooltip": "Keep model in VRAM after inference"
+      },
       "extra_config": {
         "name": "extra_config",
         "tooltip": "Extra configuration for the SAM3 model"
+      },
+      "positive_coords": {
+        "name": "positive_coords",
+        "tooltip": "Positive click coordinates as JSON: '[{\"x\": 50, \"y\": 120}]'"
+      },
+      "negative_coords": {
+        "name": "negative_coords",
+        "tooltip": "Negative click coordinates as JSON: '[{\"x\": 150, \"y\": 300}]'"
+      },
+      "bbox": {
+        "name": "bbox",
+        "tooltip": "Bounding box as (x_min, y_min, x_max, y_max) or (x, y, width, height) tuple. Compatible with KJNodes Points Editor bbox output."
       }
     },
     "outputs": {
@@ -128,6 +168,10 @@
       "1": {
         "name": "session_id",
         "tooltip": "Session ID for resuming tracking"
+      },
+      "2": {
+        "name": "objects",
+        "tooltip": "Tracked objects output data"
       }
     }
   },
@@ -174,6 +218,10 @@
         "name": "det_nms_thresh",
         "tooltip": "IoU threshold for detection NMS (Non-Maximum Suppression)"
       },
+      "new_det_thresh": {
+        "name": "new_det_thresh",
+        "tooltip": "Threshold for a detection to be added as a new object"
+      },
       "suppress_overlap_occlusion_thresh": {
         "name": "suppress_overlap_occlusion_thresh",
         "tooltip": "Threshold for suppressing overlapping objects based on recent occlusion (0.0 to disable)"
@@ -209,6 +257,10 @@
       "masklet_confirm_frames": {
         "name": "masklet_confirm_frames",
         "tooltip": "Frames needed for masklet confirmation"
+      },
+      "image_size": {
+        "name": "image_size",
+        "tooltip": "Input image size for the model"
       }
     },
     "outputs": {
diff --git a/locales/zh/nodeDefs.json b/locales/zh/nodeDefs.json
index 00d6aa2..17d7b79 100644
--- a/locales/zh/nodeDefs.json
+++ b/locales/zh/nodeDefs.json
@@ -51,6 +51,22 @@
       "add_background": {
         "name": "添加背景",
         "tooltip": "为分割的图像添加背景颜色"
+      },
+      "coordinates_positive": {
+        "name": "正向坐标",
+        "tooltip": "用于细化的正向点击坐标"
+      },
+      "coordinates_negative": {
+        "name": "负向坐标",
+        "tooltip": "用于细化的负向点击坐标"
+      },
+      "bboxes": {
+        "name": "边界框",
+        "tooltip": "用于对象检测的边界框"
+      },
+      "mask": {
+        "name": "遮罩",
+        "tooltip": "用于细化的输入遮罩"
       }
     },
     "outputs": {
@@ -61,6 +77,14 @@
       "1": {
         "name": "图像",
         "tooltip": "分割图像"
+      },
+      "2": {
+        "name": "边界框",
+        "tooltip": "检测到的边界框"
+      },
+      "3": {
+        "name": "分数",
+        "tooltip": "检测置信度分数"
       }
     }
   },
@@ -83,6 +107,14 @@
         "name": "提示词",
         "tooltip": "要跟踪的对象的文本描述（例如：'人'、'汽车'）"
       },
+      "frame_index": {
+        "name": "帧索引",
+        "tooltip": "应用初始提示词的帧"
+      },
+      "object_id": {
+        "name": "对象 ID",
+        "tooltip": "多对象跟踪的唯一 ID"
+      },
       "score_threshold_detection": {
         "name": "检测分数阈值",
         "tooltip": "检测的置信度阈值，默认为 0.5"
@@ -91,33 +123,41 @@
         "name": "新检测阈值",
         "tooltip": "将检测添加为新对象的阈值，默认为 0.7"
       },
-      "object_frame_index": {
-        "name": "对象帧索引",
-        "tooltip": "开始跟踪的对象帧索引"
-      },
-      "object_id": {
-        "name": "对象 ID",
-        "tooltip": "要跟踪的对象 ID（1-100）"
-      },
-      "start_to_propagate": {
-        "name": "开始传播",
-        "tooltip": "传播方向：禁用、双向、前向或后向"
+      "propagation_direction": {
+        "name": "传播方向",
+        "tooltip": "传播方向：双向、前向或后向"
       },
       "start_frame_index": {
         "name": "起始帧索引",
         "tooltip": "开始传播的帧索引"
       },
-      "keep_model_loaded": {
-        "name": "保持模型加载",
-        "tooltip": "推理后将模型保留在显存中"
+      "max_frames_to_track": {
+        "name": "最大跟踪帧数",
+        "tooltip": "高级：要处理的最大帧数（-1 表示全部）"
       },
       "close_after_propagation": {
         "name": "传播后关闭",
         "tooltip": "传播后关闭会话"
       },
+      "keep_model_loaded": {
+        "name": "保持模型加载",
+        "tooltip": "推理后将模型保留在显存中"
+      },
       "extra_config": {
         "name": "额外配置",
         "tooltip": "SAM3 模型的额外配置"
+      },
+      "positive_coords": {
+        "name": "正向坐标",
+        "tooltip": "正向点击坐标，JSON 格式：'[{\"x\": 50, \"y\": 120}]'"
+      },
+      "negative_coords": {
+        "name": "负向坐标",
+        "tooltip": "负向点击坐标，JSON 格式：'[{\"x\": 150, \"y\": 300}]'"
+      },
+      "bbox": {
+        "name": "边界框",
+        "tooltip": "边界框，格式为 (x_min, y_min, x_max, y_max) 或 (x, y, width, height) 元组。兼容 KJNodes Points Editor bbox 输出。"
       }
     },
     "outputs": {
@@ -128,6 +168,10 @@
       "1": {
         "name": "会话 ID",
         "tooltip": "用于恢复跟踪的会话 ID"
+      },
+      "2": {
+        "name": "对象",
+        "tooltip": "跟踪对象输出数据"
       }
     }
   },
@@ -174,6 +218,10 @@
         "name": "检测 NMS 阈值",
         "tooltip": "检测 NMS（非极大值抑制）的 IoU 阈值"
       },
+      "new_det_thresh": {
+        "name": "新检测阈值",
+        "tooltip": "将检测添加为新对象的阈值"
+      },
       "suppress_overlap_occlusion_thresh": {
         "name": "抑制重叠遮挡阈值",
         "tooltip": "基于最近遮挡抑制重叠对象的阈值（0.0 禁用）"
@@ -209,6 +257,10 @@
       "masklet_confirm_frames": {
         "name": "掩码确认帧数",
         "tooltip": "掩码确认所需的帧数"
+      },
+      "image_size": {
+        "name": "图像尺寸",
+        "tooltip": "模型的输入图像尺寸"
       }
     },
     "outputs": {
diff --git a/nodes.py b/nodes.py
index d2be316..0919db0 100644
--- a/nodes.py
+++ b/nodes.py
@@ -13,7 +13,7 @@ from PIL import Image
 from typing import Tuple, Any
 from comfy_api.latest import ComfyExtension, io
 from .sam3.logger import get_logger
-from .utils import tensor_to_pil, pil_to_tensor, masks_to_tensor, join_image_with_alpha
+from .utils import tensor_to_pil, pil_to_tensor, masks_to_tensor, join_image_with_alpha, parse_points, parse_bbox
 
 logger = get_logger(__name__)
 
@@ -40,6 +40,7 @@ class LoadSam3Model(io.ComfyNode):
                 io.Combo.Input(
                     "model",
                     options=folder_paths.get_filename_list("sam3"),
+                    default="sam3.pt",
                     tooltip="Select SAM3 model file to load"
                 ),
                 io.Combo.Input(
@@ -78,6 +79,9 @@ class LoadSam3Model(io.ComfyNode):
         if model_path is None:
             raise ValueError(f"Model file '{model}' not found in sam3 folder")
 
+        if "fp16" in model.lower():
+            precision = "fp16"
+
         # Build model based on segmentor type
         if segmentor == "image":
             from .sam3.model.sam3_image_processor import Sam3Processor
@@ -107,6 +111,14 @@ class LoadSam3Model(io.ComfyNode):
 
         logger.info("Sam3 Model loaded successfully")
 
+        if precision != 'fp32' and device == 'cpu':
+            raise ValueError("fp16 and bf16 are not supported on cpu")
+
+        if device == "cuda":
+            if torch.cuda.get_device_properties(0).major >= 8:
+                # turn on tfloat32 for Ampere GPUs (https://pytorch.org/docs/stable/notes/cuda.html#tensorfloat-32-tf32-on-ampere-devices)
+                torch.backends.cuda.matmul.allow_tf32 = True
+                torch.backends.cudnn.allow_tf32 = True
         dtype = {"bf16": torch.bfloat16, "fp16": torch.float16, "fp32": torch.float32}[precision]
         device = {"cuda": torch.device("cuda"), "cpu": torch.device("cpu"), "mps": torch.device("mps")}[device]
 
@@ -149,7 +161,7 @@ class Sam3ImageSegmentation(io.ComfyNode):
                 ),
                 io.Float.Input(
                     "threshold",
-                    default=0.60,
+                    default=0.40,
                     min=0.0,
                     max=1.0,
                     step=0.05,
@@ -164,33 +176,29 @@ class Sam3ImageSegmentation(io.ComfyNode):
                     options=["none", "black", "white", "grey"],
                     default="none",
                     tooltip="Add background color to segmented images"
-                )
-                # io.String.Input(
-                #     "coordinates_positive",
-                #     display_name="coordinates_positive",
-                #     optional=True,
-                #     force_input=True,
-                # ),
-                # io.String.Input(
-                #     "coordinates_negative",
-                #     display_name="coordinates_negative",
-                #     optional=True,
-                #     force_input=True,
-                # ),
-                # io.BBOX.Input(
-                #     "bboxes",
-                #     display_name="bboxes",
-                #     optional=True,
-                # ),
-                # io.Mask.Input(
-                #     "mask",
-                #     display_name="mask",
-                #     optional=True,
-                # ),
-                # io.Boolean.Input(
-                #     "enable_visualize",
-                #     default=False,
-                # ),
+                ),
+                io.String.Input(
+                    "coordinates_positive",
+                    display_name="coordinates_positive",
+                    optional=True,
+                    force_input=True,
+                ),
+                io.String.Input(
+                    "coordinates_negative",
+                    display_name="coordinates_negative",
+                    optional=True,
+                    force_input=True,
+                ),
+                io.BBOX.Input(
+                    "bboxes",
+                    display_name="bboxes",
+                    optional=True,
+                ),
+                io.Mask.Input(
+                    "mask",
+                    display_name="mask",
+                    optional=True,
+                ),
             ],
             outputs=[
                 io.Mask.Output(
@@ -205,12 +213,16 @@ class Sam3ImageSegmentation(io.ComfyNode):
                     is_output_list=True,
                     tooltip="Segmentation images",
                 ),
-                # io.Image.Output(
-                #     "visualization",
-                #     display_name="visualization",
-                #     is_output_list=True,
-                #     tooltip="When enable_visualize is True, the visualized image is output, otherwise the original image is output.",
-                # )
+                io.String.Output(
+                    "boxes",
+                    display_name="boxes",
+                    is_output_list=True,
+                ),
+                io.String.Output(
+                    "scores",
+                    display_name="scores",
+                    is_output_list=True,
+                ),
             ]
         )
 
@@ -232,56 +244,36 @@ class Sam3ImageSegmentation(io.ComfyNode):
         # set confidence threshold
         processor.set_confidence_threshold(threshold)
 
-        # Todo: support for points and bbox prompts
-        
-        # # handle point coordinates
-        # if coordinates_positive is not None:
-        #     try:
-        #         coordinates_positive = json.loads(coordinates_positive.replace("'", '"'))
-        #         coordinates_positive = [(coord['x'], coord['y']) for coord in coordinates_positive]
-        #         if coordinates_negative is not None:
-        #             coordinates_negative = json.loads(coordinates_negative.replace("'", '"'))
-        #             coordinates_negative = [(coord['x'], coord['y']) for coord in coordinates_negative]
-        #     except:
-        #         pass
+        # Parse inputs with bounds checking
+        pos_points, pos_count, pos_errors = parse_points(coordinates_positive, images.shape)
+        neg_points, neg_count, neg_errors = parse_points(coordinates_negative, images.shape)
+        # Combine points for refinement
+        points = None
+        point_labels = None
+        if pos_points is not None and neg_points is not None:
+            points = pos_points + neg_points
+            point_labels = [1] * pos_count + [0] * neg_count
+        elif pos_points is not None:
+            points = pos_points
+            point_labels = [1] * pos_count
+        elif neg_points is not None:
+            points = neg_points
+            point_labels = [0] * neg_count
 
-        #     positive_point_coords = np.atleast_2d(np.array(coordinates_positive))
+        # bbox
+        bounding_boxes = None
+        bounding_box_labels = None
+        if bboxes is not None:
+            bbox_coords, bbox_count = parse_bbox(bboxes, images.shape)
+            if bbox_coords is not None:
+                bounding_boxes = bbox_coords
+                bounding_box_labels = [1] * bbox_count
 
-        #     if coordinates_negative is not None:
-        #         negative_point_coords = np.array(coordinates_negative)
-        #         # Ensure both positive and negative coords are lists of 2D arrays if individual_objects is True
-        #         final_coords = np.concatenate((positive_point_coords, negative_point_coords), axis=0)
-        #     else:
-        #         final_coords = positive_point_coords
-
-        # # Handle possible bboxes
-        # if bboxes is not None:
-        #     boxes_np_batch = []
-        #     for bbox_list in bboxes:
-        #         boxes_np = []
-        #         for bbox in bbox_list:
-        #             boxes_np.append(bbox)
-        #         boxes_np = np.array(boxes_np)
-        #         boxes_np_batch.append(boxes_np)
-        #     final_box = np.array(boxes_np)
-        #     final_labels = None
-
-        # # handle labels
-        # if coordinates_positive is not None:
-        #     positive_point_labels = np.ones(len(positive_point_coords))
-
-        #     if coordinates_negative is not None:
-        #         negative_point_labels = np.zeros(len(negative_point_coords))  # 0 = negative
-        #         final_labels = np.concatenate((positive_point_labels, negative_point_labels), axis=0)
-        #     else:
-        #         final_labels = positive_point_labels
-        #     print("combined labels: ", final_labels)
-        #     print("combined labels shape: ", final_labels.shape)
-
-        # mask_list = []
 
         # Switch model to main device
         model.to(device)
+        if mask is not None:
+            mask.to(device)
 
         autocast_condition = not mm.is_device_mps(device)
         with torch.autocast(mm.get_autocast_device(device), dtype=dtype) if autocast_condition else nullcontext():
@@ -293,6 +285,8 @@ class Sam3ImageSegmentation(io.ComfyNode):
             # Process each image separately to maintain correspondence
             output_masks = []
             output_images = []
+            output_boxes = []
+            output_scores = []
 
             # Initialize progress bar
 
@@ -308,7 +302,15 @@ class Sam3ImageSegmentation(io.ComfyNode):
                 if prompt_text:
                     state = processor.set_text_prompt(prompt_text, state)
 
-                # TODO: Add support for points and bbox prompts
+                # points
+                if points is not None and len(points) > 0:
+                    state = processor.add_point_prompt(points, point_labels, state)
+                # bbox
+                if bounding_boxes is not None and len(bounding_boxes) > 0:
+                    state = processor.add_multiple_box_prompts(bounding_boxes, bounding_box_labels, state)
+                # mask
+                if mask is not None:
+                    state = processor.add_mask_prompt(mask, state)
 
                 # Get the masks and scores for this image
                 masks = state.get('masks', None)
@@ -345,7 +347,7 @@ class Sam3ImageSegmentation(io.ComfyNode):
                 img_tensor = pil_to_tensor(pil_img)
                 mask_tensor = combined_mask.unsqueeze(0)
                 rgba_image, = join_image_with_alpha(img_tensor, mask_tensor, False)
-                
+
                 if add_background != "none":
                     if add_background == "black":
                         bg_color = torch.zeros_like(rgba_image[:, :, :, :3])
@@ -353,26 +355,17 @@ class Sam3ImageSegmentation(io.ComfyNode):
                         bg_color = torch.ones_like(rgba_image[:, :, :, :3])
                     elif add_background == "grey":
                         bg_color = torch.ones_like(rgba_image[:, :, :, :3]) * 0.5
-                    
+
                     rgb = rgba_image[:, :, :, :3]
                     alpha = rgba_image[:, :, :, 3:4]
-                    
+
                     composited = rgb * alpha + bg_color * (1 - alpha)
                     output_images.append([composited.squeeze(0)])
                 else:
                     output_images.append([rgba_image.squeeze(0)])
 
-                # Visualization: overlay mask on original image with color
-                # if enable_visualize:
-                #     vis_image = visualize_masks_on_image(
-                #         pil_img,
-                #         combined_mask,
-                #         boxes,
-                #         scores,
-                #         alpha=0.5
-                #     )
-                #     vis_tensor = pil_to_tensor(vis_image)
-                #     output_visualizations.append(vis_tensor)
+                output_boxes.append(boxes)
+                output_scores.append(scores)
 
                 # Update progress bar
                 processed_frames += 1
@@ -386,7 +379,7 @@ class Sam3ImageSegmentation(io.ComfyNode):
                 model.to(offload_device)
                 mm.soft_empty_cache()
 
-        return io.NodeOutput(output_masks, output_images)
+        return io.NodeOutput(output_masks, output_images, output_boxes, output_scores)
 
 
 class Sam3VideoSegmentation(io.ComfyNode):
@@ -425,6 +418,15 @@ class Sam3VideoSegmentation(io.ComfyNode):
                     min=0,
                     max=10 ** 5,
                     step=1,
+                    tooltip="Frame where initial prompt is applied",
+                ),
+                io.Int.Input(
+                    "object_id",
+                    default=1,
+                    min=1,
+                    max=1000,
+                    step=1,
+                    tooltip="Unique ID for multi-object tracking"
                 ),
                 io.Float.Input(
                     "score_threshold_detection",
@@ -454,6 +456,12 @@ class Sam3VideoSegmentation(io.ComfyNode):
                     max=10**5,
                     step=1,
                 ),
+                io.Int.Input(
+                    "max_frames_to_track",
+                    default=-1,
+                    min=-1,
+                    tooltip="Advanced: Max frames to process (-1 for all)"
+                ),
                 io.Boolean.Input(
                     "close_after_propagation",
                     default=True,
@@ -469,28 +477,26 @@ class Sam3VideoSegmentation(io.ComfyNode):
                     tooltip="Extra configuration for the SAM3 model",
                     optional=True,
                 ),
-                # io.String.Input(
-                #     "coordinates_positive",
-                #     display_name="coordinates_positive",
-                #     optional=True,
-                #     force_input=True,
-                # ),
-                # io.String.Input(
-                #     "coordinates_negative",
-                #     display_name="coordinates_negative",
-                #     optional=True,
-                #     force_input=True,
-                # ),
-                # io.BBOX.Input(
-                #     "bboxes",
-                #     display_name="bboxes",
-                #     optional=True,
-                # ),
-                # io.Mask.Input(
-                #     "mask",
-                #     display_name="mask",
-                #     optional=True,
-                # )
+                io.String.Input(
+                  "positive_coords",
+                  display_name="positive_coords",
+                  tooltip="Positive click coordinates as JSON: '[{\"x\": 50, \"y\": 120}]'",
+                  optional=True,
+                  force_input=True,
+                ),
+                io.String.Input(
+                    "negative_coords",
+                    display_name="negative_coords",
+                    tooltip="Negative click coordinates as JSON: '[{\"x\": 150, \"y\": 300}]'",
+                    optional=True,
+                    force_input=True,
+                ),
+                io.BBOX.Input(
+                    "bbox",
+                    display_name="bbox",
+                    optional=True,
+                    tooltip="Bounding box as (x_min, y_min, x_max, y_max) or (x, y, width, height) tuple. Compatible with KJNodes Points Editor bbox output."
+                ),
             ],
             outputs=[
                 io.Mask.Output(
@@ -511,8 +517,8 @@ class Sam3VideoSegmentation(io.ComfyNode):
 
 
     @classmethod
-    def execute(cls, sam3_model, video_frames, prompt, frame_index, score_threshold_detection, new_det_thresh, propagation_direction, start_frame_index=0,close_after_propagation=True,  keep_model_loaded=False, session_id=None, extra_config=None,coordinates_positive=None, coordinates_negative=None,
-                 bboxes=None, mask=None) -> io.NodeOutput:
+    def execute(cls, sam3_model, video_frames, prompt, frame_index, object_id, score_threshold_detection, new_det_thresh, propagation_direction, start_frame_index=0, max_frames_to_track=-1, close_after_propagation=True,  keep_model_loaded=False, session_id=None, extra_config=None, positive_coords=None, negative_coords=None,
+                 bbox=None,) -> io.NodeOutput:
         offload_device = mm.unet_offload_device()
 
         video_predictor = sam3_model.get("model", None)
@@ -524,6 +530,10 @@ class Sam3VideoSegmentation(io.ComfyNode):
         if video_predictor is None or segmentor != "video":
             raise ValueError("Invalid SAM3 model. Please load a SAM3 model in 'video' mode")
 
+        if frame_index > B - 1:
+            logger.info(f"Frame index {frame_index} is out of bounds, setting to last frame {B - 1}")
+            frame_index = B - 1
+
         # Set video model config
         video_predictor.model.score_threshold_detection = score_threshold_detection
         video_predictor.model.new_det_thresh = new_det_thresh
@@ -574,14 +584,46 @@ class Sam3VideoSegmentation(io.ComfyNode):
         video_predictor.model.to(device)
 
         autocast_condition = not mm.is_device_mps(device)
-        with torch.autocast(mm.get_autocast_device(device), dtype=dtype) if autocast_condition else nullcontext(): 
+        with torch.autocast(mm.get_autocast_device(device), dtype=dtype) if autocast_condition else nullcontext():
+
+            # Parse inputs with bounds checking
+            pos_points, pos_count, pos_errors = parse_points(positive_coords, video_frames.shape)
+            neg_points, neg_count, neg_errors = parse_points(negative_coords, video_frames.shape)
+            # Combine points for refinement
+            points = None
+            point_labels = None
+            if pos_points is not None and neg_points is not None:
+                points = pos_points + neg_points
+                point_labels = [1] * pos_count + [0] * neg_count
+            elif pos_points is not None:
+                points = pos_points
+                point_labels = [1] * pos_count
+            elif neg_points is not None:
+                points = neg_points
+                point_labels = [0] * neg_count
+
+            # bbox (has bugs)
+            bounding_boxes = None
+            bounding_box_labels = None
+            if bbox is not None:
+                bbox_coords, bbox_count = parse_bbox(bbox, video_frames.shape)
+                if bbox_coords is not None:
+                    bounding_boxes = bbox_coords
+                    bounding_box_labels = [1] * bbox_count
+
+            # print('bbox_coords:', bbox_coords)
             # Add Prompt
             response = video_predictor.handle_request(
                 request=dict(
                     type="add_prompt",
                     session_id=session_id,
                     frame_index=frame_index,
-                    text=prompt,
+                    text=prompt if prompt else None,
+                    bounding_boxes=bounding_boxes,
+                    bounding_box_labels=bounding_box_labels,
+                    points=points,
+                    point_labels=point_labels,
+                    obj_id=object_id
                 )
             )
 
@@ -603,6 +645,7 @@ class Sam3VideoSegmentation(io.ComfyNode):
                     session_id=session_id,
                     propagation_direction=propagation_direction,
                     start_frame_index=start_frame_index,
+                    max_frame_num_to_track=max_frames_to_track if max_frames_to_track != -1 else None,
                 )
             ):
                 frame_idx = response.get("frame_index", 0)
diff --git a/sam3/model/sam3_image_processor.py b/sam3/model/sam3_image_processor.py
index f02b08a..668bace 100644
--- a/sam3/model/sam3_image_processor.py
+++ b/sam3/model/sam3_image_processor.py
@@ -151,6 +151,90 @@ class Sam3Processor:
 
         return self._forward_grounding(state)
 
+    @torch.inference_mode()
+    def add_multiple_box_prompts(self, boxes: List[List], labels: List[bool], state: Dict):
+        """Adds multiple box prompts and run the inference.
+        The image needs to be set, but not necessarily the text prompt.
+        Each box is assumed to be in [center_x, center_y, width, height] format and normalized in [0, 1] range.
+        Each label is True for a positive box, False for a negative box.
+        """
+        if "backbone_out" not in state:
+            raise ValueError("You must call set_image before add_multiple_box_prompts")
+
+        if "language_features" not in state["backbone_out"]:
+            dummy_text_outputs = self.model.backbone.forward_text(
+                ["visual"], device=self.device
+            )
+            state["backbone_out"].update(dummy_text_outputs)
+
+        if "geometric_prompt" not in state:
+            state["geometric_prompt"] = self.model._get_dummy_prompt()
+
+        # Convert to [seq_len, batch_size, 4] format
+        boxes_tensor = torch.tensor(boxes, device=self.device, dtype=torch.float32).view(len(boxes), 1, 4)
+        labels_tensor = torch.tensor(labels, device=self.device, dtype=torch.bool).view(len(labels), 1)
+        state["geometric_prompt"].append_boxes(boxes_tensor, labels_tensor)
+
+        return self._forward_grounding(state)
+
+    @torch.inference_mode()
+    def add_point_prompt(self, points: List[List], labels: List[int], state: Dict):
+        """Adds point prompts and run the inference.
+        The image needs to be set, but not necessarily the text prompt.
+        Points should be in [x, y] format, normalized in [0, 1] range.
+        Labels should be 1 for foreground points, 0 for background points.
+        """
+        if "backbone_out" not in state:
+            raise ValueError("You must call set_image before add_point_prompt")
+
+        if "language_features" not in state["backbone_out"]:
+            dummy_text_outputs = self.model.backbone.forward_text(
+                ["visual"], device=self.device
+            )
+            state["backbone_out"].update(dummy_text_outputs)
+
+        if "geometric_prompt" not in state:
+            state["geometric_prompt"] = self.model._get_dummy_prompt()
+
+        # Convert to [seq_len, batch_size, 2] format
+        points_tensor = torch.tensor(points, device=self.device, dtype=torch.float32).view(len(points), 1, 2)
+        labels_tensor = torch.tensor(labels, device=self.device, dtype=torch.long).view(len(labels), 1)
+        state["geometric_prompt"].append_points(points_tensor, labels_tensor)
+
+        return self._forward_grounding(state)
+
+    @torch.inference_mode()
+    def add_mask_prompt(self, mask: torch.Tensor, state: Dict):
+        """Adds a mask prompt and run the inference.
+        The mask should be a binary tensor with shape matching the model's expected input.
+        This is typically used for iterative refinement.
+        """
+        if "backbone_out" not in state:
+            raise ValueError("You must call set_image before add_mask_prompt")
+
+        if "language_features" not in state["backbone_out"]:
+            dummy_text_outputs = self.model.backbone.forward_text(
+                ["visual"], device=self.device
+            )
+            state["backbone_out"].update(dummy_text_outputs)
+
+        if "geometric_prompt" not in state:
+            state["geometric_prompt"] = self.model._get_dummy_prompt()
+
+        # Ensure mask is on correct device and has batch dimension
+        if mask.device != self.device:
+            mask = mask.to(self.device)
+
+        # Add sequence and batch dimensions if needed: [seq_len, batch_size, H, W]
+        if len(mask.shape) == 2:  # [H, W]
+            mask = mask.unsqueeze(0).unsqueeze(0)
+        elif len(mask.shape) == 3:  # [1, H, W] or [batch, H, W]
+            mask = mask.unsqueeze(0)
+
+        state["geometric_prompt"].append_masks(mask)
+
+        return self._forward_grounding(state)
+
     def reset_all_prompts(self, state: Dict):
         """Removes all the prompts and results"""
         if "backbone_out" in state:
diff --git a/sam3/model/sam3_tracking_predictor.py b/sam3/model/sam3_tracking_predictor.py
index 6563320..db1f807 100644
--- a/sam3/model/sam3_tracking_predictor.py
+++ b/sam3/model/sam3_tracking_predictor.py
@@ -482,7 +482,7 @@ class Sam3TrackerPredictor(Sam3TrackerBase):
         if self.non_overlap_masks_for_output:
             video_res_masks = self._apply_non_overlapping_constraints(video_res_masks)
         # potentially fill holes in the predicted masks
-        if self.fill_hole_area > 0:
+        if self.fill_hole_area > 0 and len(video_res_masks) > 0:
             video_res_masks = fill_holes_in_mask_scores(
                 video_res_masks, self.fill_hole_area
             )
diff --git a/sam3/model/sam3_video_base.py b/sam3/model/sam3_video_base.py
index e74faaa..b870dab 100644
--- a/sam3/model/sam3_video_base.py
+++ b/sam3/model/sam3_video_base.py
@@ -966,12 +966,13 @@ class Sam3VideoBase(nn.Module):
         # Part 2: masks from new detections
         new_det_fa_inds_t = torch.from_numpy(new_det_fa_inds)
         new_det_low_res_masks = det_out["mask"][new_det_fa_inds_t].unsqueeze(1)
-        new_det_low_res_masks = fill_holes_in_mask_scores(
-            new_det_low_res_masks,
-            max_area=self.fill_hole_area,
-            fill_holes=True,
-            remove_sprinkles=True,
-        )
+        if len(new_det_fa_inds) > 0:
+            new_det_low_res_masks = fill_holes_in_mask_scores(
+                new_det_low_res_masks,
+                max_area=self.fill_hole_area,
+                fill_holes=True,
+                remove_sprinkles=True,
+            )
         new_masklet_video_res_masks = F.interpolate(
             new_det_low_res_masks,
             size=(orig_vid_height, orig_vid_width),
diff --git a/sam3/model/sam3_video_inference.py b/sam3/model/sam3_video_inference.py
index ac6af8e..921dd3f 100644
--- a/sam3/model/sam3_video_inference.py
+++ b/sam3/model/sam3_video_inference.py
@@ -84,7 +84,7 @@ class Sam3VideoInference(Sam3VideoBase):
         inference_state["feature_cache"] = {}
         inference_state["cached_frame_outputs"] = {}
         inference_state["action_history"] = []  # for logging user actions
-        inference_state["is_image_only"] = is_image_type(resource_path) if resource_path else False
+        inference_state["is_image_only"] = is_image_type(resource_path)
         return inference_state
 
     @torch.inference_mode()
@@ -1154,6 +1154,12 @@ class Sam3VideoInferenceWithInstanceInteractivity(Sam3VideoInference):
                         )  # (1, H_video, W_video) bool
                         refined_obj_id_to_mask[obj_id] = refined_mask_video_res
 
+                    # Initialize cache if not present (needed for point prompts during propagation)
+                    if "cached_frame_outputs" not in inference_state:
+                        inference_state["cached_frame_outputs"] = {}
+                    if frame_idx not in inference_state["cached_frame_outputs"]:
+                        inference_state["cached_frame_outputs"][frame_idx] = {}
+
                     obj_id_to_mask = self._build_tracker_output(
                         inference_state, frame_idx, refined_obj_id_to_mask
                     )
@@ -1575,6 +1581,12 @@ class Sam3VideoInferenceWithInstanceInteractivity(Sam3VideoInference):
             new_mask_data = data_list[0].to(self.device)
 
         if self.rank == 0:
+            # Initialize cache if not present (needed for point prompts without prior propagation)
+            if "cached_frame_outputs" not in inference_state:
+                inference_state["cached_frame_outputs"] = {}
+            if frame_idx not in inference_state["cached_frame_outputs"]:
+                inference_state["cached_frame_outputs"][frame_idx] = {}
+
             obj_id_to_mask = self._build_tracker_output(
                 inference_state,
                 frame_idx,
@@ -1706,4 +1718,4 @@ class Sam3VideoInferenceWithInstanceInteractivity(Sam3VideoInference):
 def is_image_type(resource_path: str) -> bool:
     if isinstance(resource_path, list):
         return len(resource_path) == 1
-    return resource_path.lower().endswith(tuple(IMAGE_EXTS))
+    return resource_path.lower().endswith(tuple(IMAGE_EXTS))
\ No newline at end of file
diff --git a/sam3/model_builder.py b/sam3/model_builder.py
index 24a52ed..b04dfec 100644
--- a/sam3/model_builder.py
+++ b/sam3/model_builder.py
@@ -309,18 +309,6 @@ def _create_sam3_model(
     }
 
     matcher = None
-    if not eval_mode:
-        from .train.matcher import BinaryHungarianMatcherV2
-
-        matcher = BinaryHungarianMatcherV2(
-            focal=True,
-            cost_class=2.0,
-            cost_bbox=5.0,
-            cost_giou=2.0,
-            alpha=0.25,
-            gamma=2,
-            stable=False,
-        )
     common_params["matcher"] = matcher
     model = Sam3Image(**common_params)
 
diff --git a/sam3/sam3_image_seg_by_prompt.json b/sam3/sam3_image_seg_by_prompt.json
new file mode 100644
index 0000000..dcc1cbf
--- /dev/null
+++ b/sam3/sam3_image_seg_by_prompt.json
@@ -0,0 +1,281 @@
+{
+  "id": "307743a2-8a4b-4090-b8dc-3125ff31fa14",
+  "revision": 0,
+  "last_node_id": 223,
+  "last_link_id": 304,
+  "nodes": [
+    {
+      "id": 222,
+      "type": "PreviewImage",
+      "pos": [
+        -3883.9588057841556,
+        593.0894010290336
+      ],
+      "size": [
+        339.2921196124935,
+        246
+      ],
+      "flags": {},
+      "order": 4,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "images",
+          "type": "IMAGE",
+          "link": 303
+        }
+      ],
+      "outputs": [],
+      "properties": {
+        "cnr_id": "comfy-core",
+        "ver": "0.3.70",
+        "Node name for S&R": "PreviewImage"
+      },
+      "widgets_values": []
+    },
+    {
+      "id": 218,
+      "type": "MaskPreview+",
+      "pos": [
+        -3880.8437110898844,
+        280.9015183807766
+      ],
+      "size": [
+        332.95740427927046,
+        258
+      ],
+      "flags": {},
+      "order": 3,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "mask",
+          "type": "MASK",
+          "link": 298
+        }
+      ],
+      "outputs": [],
+      "properties": {
+        "cnr_id": "comfyui_essentials",
+        "ver": "9d9f4bedfc9f0321c19faf71855e228c93bd0dc9",
+        "Node name for S&R": "MaskPreview+"
+      },
+      "widgets_values": []
+    },
+    {
+      "id": 190,
+      "type": "easy sam3ModelLoader",
+      "pos": [
+        -4235.041825672723,
+        334.7519653807837
+      ],
+      "size": [
+        314.09852906952574,
+        130
+      ],
+      "flags": {},
+      "order": 0,
+      "mode": 0,
+      "inputs": [],
+      "outputs": [
+        {
+          "name": "sam3_model",
+          "type": "EASY_SAM3_MODEL",
+          "links": [
+            282
+          ]
+        }
+      ],
+      "properties": {
+        "Node name for S&R": "easy sam3ModelLoader"
+      },
+      "widgets_values": [
+        "sam3.pt",
+        "image",
+        "cuda",
+        "fp16"
+      ]
+    },
+    {
+      "id": 221,
+      "type": "LoadImage",
+      "pos": [
+        -4678.319118164902,
+        372.4935004943305
+      ],
+      "size": [
+        406.60579628243977,
+        446.341753250938
+      ],
+      "flags": {},
+      "order": 1,
+      "mode": 0,
+      "inputs": [],
+      "outputs": [
+        {
+          "name": "IMAGE",
+          "type": "IMAGE",
+          "links": [
+            302
+          ]
+        },
+        {
+          "name": "MASK",
+          "type": "MASK",
+          "links": null
+        }
+      ],
+      "properties": {
+        "cnr_id": "comfy-core",
+        "ver": "0.3.70",
+        "Node name for S&R": "LoadImage"
+      },
+      "widgets_values": [
+        "1 (3).png",
+        "image"
+      ]
+    },
+    {
+      "id": 213,
+      "type": "easy sam3ImageSegmentation",
+      "pos": [
+        -4241.387303264146,
+        538.9760914950197
+      ],
+      "size": [
+        337.2438563422779,
+        281.966544722883
+      ],
+      "flags": {},
+      "order": 2,
+      "mode": 0,
+      "inputs": [
+        {
+          "name": "sam3_model",
+          "type": "EASY_SAM3_MODEL",
+          "link": 282
+        },
+        {
+          "name": "images",
+          "type": "IMAGE",
+          "link": 302
+        },
+        {
+          "name": "coordinates_positive",
+          "shape": 7,
+          "type": "STRING",
+          "link": null
+        },
+        {
+          "name": "coordinates_negative",
+          "shape": 7,
+          "type": "STRING",
+          "link": null
+        },
+        {
+          "name": "bboxes",
+          "shape": 7,
+          "type": "BBOX",
+          "link": null
+        },
+        {
+          "name": "mask",
+          "shape": 7,
+          "type": "MASK",
+          "link": null
+        }
+      ],
+      "outputs": [
+        {
+          "name": "masks",
+          "shape": 6,
+          "type": "MASK",
+          "links": [
+            298
+          ]
+        },
+        {
+          "name": "images",
+          "shape": 6,
+          "type": "IMAGE",
+          "links": [
+            303
+          ]
+        },
+        {
+          "name": "boxes",
+          "shape": 6,
+          "type": "STRING",
+          "links": null
+        },
+        {
+          "name": "scores",
+          "shape": 6,
+          "type": "STRING",
+          "links": null
+        }
+      ],
+      "properties": {
+        "Node name for S&R": "easy sam3ImageSegmentation"
+      },
+      "widgets_values": [
+        "apple",
+        0.4,
+        false,
+        "none"
+      ]
+    }
+  ],
+  "links": [
+    [
+      282,
+      190,
+      0,
+      213,
+      0,
+      "EASY_SAM3_MODEL"
+    ],
+    [
+      298,
+      213,
+      0,
+      218,
+      0,
+      "MASK"
+    ],
+    [
+      302,
+      221,
+      0,
+      213,
+      1,
+      "IMAGE"
+    ],
+    [
+      303,
+      213,
+      1,
+      222,
+      0,
+      "IMAGE"
+    ]
+  ],
+  "groups": [],
+  "config": {},
+  "extra": {
+    "ds": {
+      "scale": 1.3920191078517683,
+      "offset": [
+        4898.859436279234,
+        -79.3314399803767
+      ]
+    },
+    "workflowRendererVersion": "LG",
+    "frontendVersion": "1.33.5",
+    "VHS_latentpreview": false,
+    "VHS_latentpreviewrate": 0,
+    "VHS_MetadataImage": true,
+    "VHS_KeepIntermediate": true
+  },
+  "version": 0.4
+}
\ No newline at end of file
diff --git a/utils.py b/utils.py
index 1803327..a4e582a 100644
--- a/utils.py
+++ b/utils.py
@@ -4,6 +4,7 @@ Utility functions for tensor and PIL image conversions.
 
 import numpy as np
 import torch
+import json
 from PIL import Image
 from typing import List, Union, Optional
 
@@ -228,4 +229,216 @@ def join_image_with_alpha(image: torch.Tensor, alpha: torch.Tensor, invert=False
     for i in range(batch_size):
         out_images.append(torch.cat((image[i][:,:,:3], alpha[i].unsqueeze(2)), dim=2))
 
-    return torch.stack(out_images),
\ No newline at end of file
+    return torch.stack(out_images),
+
+def parse_points(points_str, image_shape=None):
+    """Parse point coordinates from JSON string and validate bounds.
+    
+    Converts pixel coordinates to normalized coordinates (0-1 range) if image_shape is provided.
+
+    Returns:
+        tuple: (points_array, labels_array, validation_errors) where validation_errors
+               is a list of error messages, or (None, None, errors) if all points invalid
+    """
+    if not points_str or not points_str.strip():
+        return None, None, []
+
+    try:
+        points_list = json.loads(points_str)
+
+        if not isinstance(points_list, list):
+            raise ValueError(f"Points must be a JSON array, got {type(points_list).__name__}")
+
+        if len(points_list) == 0:
+            return None, None, []
+
+        points = []
+        validation_errors = []
+
+        for i, point_dict in enumerate(points_list):
+            if not isinstance(point_dict, dict):
+                err = f"Point {i} is not a dictionary"
+                print(f"Warning: {err}, skipping")
+                validation_errors.append(err)
+                continue
+
+            if 'x' not in point_dict or 'y' not in point_dict:
+                err = f"Point {i} missing 'x' or 'y' key"
+                print(f"Warning: {err}, skipping")
+                validation_errors.append(err)
+                continue
+
+            try:
+                x = float(point_dict['x'])
+                y = float(point_dict['y'])
+
+                # Validate coordinates are non-negative
+                if x < 0 or y < 0:
+                    err = f"Point {i} has negative coordinates ({x}, {y})"
+                    print(f"Warning: {err}, skipping")
+                    validation_errors.append(err)
+                    continue
+
+                # Normalize to 0-1 range if image shape is provided
+                if image_shape is not None:
+                    height, width = image_shape[1], image_shape[2]  # [batch, height, width, channels]
+                    
+                    # Validate within image bounds
+                    if x >= width or y >= height:
+                        err = f"Point {i} ({x}, {y}) outside image bounds ({width}x{height})"
+                        print(f"Warning: {err}, skipping")
+                        validation_errors.append(err)
+                        continue
+                    
+                    # Normalize coordinates to [0, 1] range
+                    x = x / width
+                    y = y / height
+
+                points.append([x, y])
+
+            except (ValueError, TypeError) as e:
+                err = f"Could not convert point {i} coordinates to float: {e}"
+                print(f"Warning: {err}, skipping")
+                validation_errors.append(err)
+                continue
+
+        if not points:
+            return None, None, validation_errors
+
+        return points, len(points), validation_errors
+
+    except json.JSONDecodeError as e:
+        raise ValueError(f"Invalid JSON in points: {str(e)}")
+    except Exception as e:
+        print(f"Error parsing points: {e}")
+        return None, None, [str(e)]
+
+def parse_bbox(bbox, image_shape=None):
+    """Parse bounding box from BBOX type (tuple/list/dict) and validate
+    
+    Converts pixel coordinates to normalized coordinates (0-1 range) if image_shape is provided.
+
+    Supports multiple formats:
+    - KJNodes: [{'startX': x, 'startY': y, 'endX': x2, 'endY': y2}, ...]
+    - Tuple/list: (x1, y1, x2, y2) or (x, y, width, height)
+    - Dict: {'startX': x, 'startY': y, 'endX': x2, 'endY': y2}
+    
+    Returns:
+        List of bounding boxes [[x1, y1, x2, y2], ...] in normalized coordinates (0-1) if image_shape provided, or None
+    """
+    if bbox is None:
+        return None
+
+    try:
+        all_coords = []
+
+        # Try to extract coordinates regardless of type checks
+        # This handles cases where ComfyUI wraps data in unexpected ways
+        if hasattr(bbox, '__iter__') and not isinstance(bbox, (str, bytes)):
+            # It's some kind of sequence
+            try:
+                bbox_list = list(bbox)
+
+                if len(bbox_list) == 0:
+                    return None
+
+                # Check if it's a list of 4 numbers (single bbox)
+                if len(bbox_list) == 4 and all(isinstance(x, (int, float)) for x in bbox_list):
+                    coords = [float(x) for x in bbox_list]
+                    all_coords.append(coords)
+                else:
+                    # Process each element as a potential bbox
+                    for elem in bbox_list:
+                        coords = None
+                        
+                        # Try to access as dict-like (KJNodes format)
+                        if hasattr(elem, '__getitem__'):
+                            try:
+                                x1 = float(elem['startX'])
+                                y1 = float(elem['startY'])
+                                x2 = float(elem['endX'])
+                                y2 = float(elem['endY'])
+                                coords = [x1, y1, x2, y2]
+                            except (KeyError, TypeError):
+                                # Not dict format, might be numeric sequence
+                                pass
+                        
+                        # If still no coords, try as numeric sequence
+                        if coords is None:
+                            if hasattr(elem, '__iter__') and not isinstance(elem, (str, bytes)):
+                                inner = list(elem)
+                                if len(inner) == 4:
+                                    coords = [float(x) for x in inner]
+                        
+                        if coords is not None:
+                            all_coords.append(coords)
+
+            except Exception as e:
+                raise ValueError(f"Failed to process bbox as sequence: {e}")
+
+        # Try single dict format
+        elif hasattr(bbox, '__getitem__'):
+            try:
+                x1 = float(bbox['startX'])
+                y1 = float(bbox['startY'])
+                x2 = float(bbox['endX'])
+                y2 = float(bbox['endY'])
+                coords = [x1, y1, x2, y2]
+                all_coords.append(coords)
+            except (KeyError, TypeError) as e:
+                raise ValueError(f"Dictionary bbox missing required keys: {e}")
+
+        else:
+            raise ValueError(f"Unsupported bbox type: {type(bbox)}")
+
+        if not all_coords:
+            raise ValueError(
+                f"Could not extract coordinates from bbox. Type: {type(bbox)}, Content: {repr(bbox)[:200]}")
+
+        # Process and validate each bbox
+        validated_coords = []
+        for coords in all_coords:
+            # Handle xywh format (convert to xyxy)
+            x1, y1, x2, y2 = coords
+            if x2 < x1 or y2 < y1:
+                # Assume xywh format: (x, y, width, height)
+                width, height = x2, y2
+                x2 = x1 + width
+                y2 = y1 + height
+                coords = [x1, y1, x2, y2]
+
+            # Validate coordinates
+            if coords[0] >= coords[2]:
+                raise ValueError(f"Invalid bbox: x1 ({coords[0]}) must be < x2 ({coords[2]})")
+            if coords[1] >= coords[3]:
+                raise ValueError(f"Invalid bbox: y1 ({coords[1]}) must be < y2 ({coords[3]})")
+            if coords[0] < 0 or coords[1] < 0:
+                raise ValueError(f"Bounding box coordinates must be non-negative, got x1={coords[0]}, y1={coords[1]}")
+            
+            # Normalize to 0-1 range if image shape is provided
+            if image_shape is not None:
+                height, width = image_shape[1], image_shape[2]  # [batch, height, width, channels]
+                
+                # Validate within image bounds
+                if coords[0] >= width or coords[2] > width:
+                    print(f"Warning: bbox x coordinates ({coords[0]}, {coords[2]}) outside image width ({width})")
+                if coords[1] >= height or coords[3] > height:
+                    print(f"Warning: bbox y coordinates ({coords[1]}, {coords[3]}) outside image height ({height})")
+                
+                # Normalize coordinates to [0, 1] range
+                coords = [
+                    coords[0] / width,   # x1
+                    coords[1] / height,  # y1
+                    coords[2] / width,   # x2
+                    coords[3] / height   # y2
+                ]
+            
+            validated_coords.append(coords)
+
+        return validated_coords, len(validated_coords)
+
+    except (ValueError, TypeError) as e:
+        error_msg = f"Invalid bbox: {str(e)}\n"
+        error_msg += f"Input type: {type(bbox)}\n"
+        error_msg += f"Input content: {repr(bbox)[:500]}"
+        raise ValueError(error_msg)
\ No newline at end of file