Update README

This commit is contained in:
yolain
2025-11-22 18:33:13 +08:00
parent 5ffffeaf68
commit 87d46bbba3
2 changed files with 4 additions and 4 deletions
+1 -1
View File
@@ -17,7 +17,7 @@ This node package brings Meta's SAM3 model to ComfyUI, enabling:
### Video Segmentation
![Video Segmentation Example](assets/video.png)
*Example of semantic segmentation on video frames*
*Example of points coordinates segmentation on video frames*
## Features
+3 -3
View File
@@ -17,7 +17,7 @@
### 视频分割
![视频分割示例](assets/video.png)
*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