Update README
This commit is contained in:
@@ -17,7 +17,7 @@ This node package brings Meta's SAM3 model to ComfyUI, enabling:
|
||||
|
||||
### Video Segmentation
|
||||

|
||||
*Example of semantic segmentation on video frames*
|
||||
*Example of points coordinates segmentation on video frames*
|
||||
|
||||
## Features
|
||||
|
||||
|
||||
+3
-3
@@ -17,7 +17,7 @@
|
||||
|
||||
### 视频分割
|
||||

|
||||
*视频帧语义分割示例*
|
||||
*视频帧点坐标分割示例*
|
||||
|
||||
|
||||
## 功能特性
|
||||
@@ -52,8 +52,8 @@
|
||||
- `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
|
||||
- `coordinates_positive`(可选):正向点坐标以细化分割
|
||||
- `coordinates_negative`(可选):负向点坐标以排除区域
|
||||
- `bboxes`(可选):边界框来引导分割
|
||||
- `mask`(可选):用于细化的输入遮罩
|
||||
|
||||
|
||||
Reference in New Issue
Block a user