suport more version BiRefNet models

README.md

@@ -9,6 +9,11 @@
 
 本仓库将BiRefNet最新模型封装为ComfyUI节点来使用，相较于旧模型，最新模型的抠图精度更高更好。<br>
 
+
+## 🔥🔥🔥 更新!!
+
+* 2025年02月17日: 🚀 支持多个版本BiRefNet模型，比如 [BiRefNet](https://huggingface.co/ZhengPeng7/BiRefNet) 、[BiRefNet_HR](https://huggingface.co/ZhengPeng7/BiRefNet_HR) 、 [BiRefNet-portrait](https://huggingface.co/ZhengPeng7/BiRefNet-portrait) 。
+
 ## 安装 
 
 #### 方法1:

README_EN.md

@@ -8,6 +8,10 @@
 
 This repository wraps the latest BiRefNet model as ComfyUI nodes. Compared to the previous model, the latest model offers higher and better matting accuracy.
 
+## 🔥🔥🔥 Update!!
+
+* 2025/02/17: 🚀 Multiple versions of BiRefNet models are supported, such as [BiRefNet](https://huggingface.co/ZhengPeng7/BiRefNet) 、[BiRefNet_HR](https://huggingface.co/ZhengPeng7/BiRefNet_HR) 、 [BiRefNet-portrait](https://huggingface.co/ZhengPeng7/BiRefNet-portrait) .
+
 ## Installation 
 
 #### Method  1:

js/BiRefNet_Hugo.js

@@ -81,8 +81,8 @@ app.registerExtension({
 				return;
 			}
 
-			if(node.widgets_values[3] != undefined){
-				localModelPathWidget.value = node.widgets_values[3];
+			if(node.widgets_values[4] != undefined){
+				localModelPathWidget.value = node.widgets_values[4];
 			}
 			
 			// 初始状态控制：根据 load_local_model 的值决定是否显示 local_model_path

nodes.py

@@ -65,6 +65,7 @@ class BiRefNet_Hugo:
         return {
             "required": {
                 "image": ("IMAGE",),
+                "model": (["ZhengPeng7/BiRefNet", "ZhengPeng7/BiRefNet_HR", "ZhengPeng7/BiRefNet-portrait"],{"default": "ZhengPeng7/BiRefNet"}),
                 "load_local_model": ("BOOLEAN", {"default": False}),
                 "background_color_name": (colors,{"default": "transparency"}), 
                 "device": (["auto", "cuda", "cpu", "mps", "xpu", "meta"],{"default": "auto"})
@@ -81,6 +82,7 @@ class BiRefNet_Hugo:
   
     def background_remove(self, 
                           image, 
+                          model,
                           load_local_model,
                           device, 
                           background_color_name, 
@@ -96,7 +98,7 @@ class BiRefNet_Hugo:
             birefnet = AutoModelForImageSegmentation.from_pretrained(local_model_path,trust_remote_code=True)
         else:
             birefnet = AutoModelForImageSegmentation.from_pretrained(
-                "ZhengPeng7/BiRefNet", trust_remote_code=True
+                model, trust_remote_code=True
             )
         
         birefnet.to(device)