clipseg improvements

__init__.py

@@ -19,6 +19,7 @@ NODE_CONFIG = {
     "ConditioningSetMaskAndCombine5": {"class": ConditioningSetMaskAndCombine5, "name": "ConditioningSetMaskAndCombine5"},
     "CondPassThrough": {"class": CondPassThrough},
     #masking
+    "DownloadAndLoadCLIPSeg": {"class": DownloadAndLoadCLIPSeg, "name": "(Down)load CLIPSeg"},
     "BatchCLIPSeg": {"class": BatchCLIPSeg, "name": "Batch CLIPSeg"},
     "ColorToMask": {"class": ColorToMask, "name": "Color To Mask"},
     "CreateGradientMask": {"class": CreateGradientMask, "name": "Create Gradient Mask"},

nodes/mask_nodes.py

@@ -31,7 +31,7 @@ class BatchCLIPSeg:
                     {
                         "images": ("IMAGE",),
                         "text": ("STRING", {"multiline": False}),
-                        "threshold": ("FLOAT", {"default": 0.1,"min": 0.0, "max": 10.0, "step": 0.001}),
+                        "threshold": ("FLOAT", {"default": 0.5,"min": 0.0, "max": 10.0, "step": 0.001}),
                         "binary_mask": ("BOOLEAN", {"default": True}),
                         "combine_mask": ("BOOLEAN", {"default": False}),
                         "use_cuda": ("BOOLEAN", {"default": True}),
@@ -39,6 +39,8 @@ class BatchCLIPSeg:
                      "optional":
                     {
                         "blur_sigma": ("FLOAT", {"default": 0.0, "min": 0.0, "max": 100.0, "step": 0.1}),
+                        "opt_model": ("CLIPSEGMODEL", ),
+                        "prev_mask": ("MASK", {"default": None}),
                     }
                 }
 
@@ -50,7 +52,7 @@ class BatchCLIPSeg:
 Segments an image or batch of images using CLIPSeg.
 """
 
-    def segment_image(self, images, text, threshold, binary_mask, combine_mask, use_cuda, blur_sigma=0.0):
+    def segment_image(self, images, text, threshold, binary_mask, combine_mask, use_cuda, blur_sigma=0.0, opt_model=None, prev_mask=None):
         from transformers import CLIPSegProcessor, CLIPSegForImageSegmentation
         import torchvision.transforms as transforms
         offload_device = model_management.unet_offload_device()
@@ -59,10 +61,23 @@ Segments an image or batch of images using CLIPSeg.
         else:
             device = torch.device("cpu")
         dtype = model_management.unet_dtype()
-        if not hasattr(self, "model"):
-            self.model = CLIPSegForImageSegmentation.from_pretrained("CIDAS/clipseg-rd64-refined")
-        
-        processor = CLIPSegProcessor.from_pretrained("CIDAS/clipseg-rd64-refined")
+
+        if opt_model is None:
+            checkpoint_path = os.path.join(folder_paths.models_dir,'clip_seg', 'clipseg-rd64-refined-fp16')
+            if not hasattr(self, "model"):
+                try:
+                    if not os.path.exists(checkpoint_path):
+                        from huggingface_hub import snapshot_download
+                        snapshot_download(repo_id="Kijai/clipseg-rd64-refined-fp16", local_dir=checkpoint_path, local_dir_use_symlinks=False)
+                    self.model = CLIPSegForImageSegmentation.from_pretrained(checkpoint_path)
+                except:
+                    checkpoint_path = "CIDAS/clipseg-rd64-refined"
+                    self.model = CLIPSegForImageSegmentation.from_pretrained(checkpoint_path)
+            processor = CLIPSegProcessor.from_pretrained(checkpoint_path)
+
+        else:
+            self.model = opt_model['model']
+            processor = opt_model['processor']
 
         self.model.to(dtype).to(device)
 
@@ -81,20 +96,20 @@ Segments an image or batch of images using CLIPSeg.
             outputs = self.model(**input_prc)
 
         tensor = torch.sigmoid(outputs.logits)
-        tensor = torch.where(tensor > (threshold / 10), tensor, torch.tensor(0, dtype=torch.float))
+        print(tensor.min(), tensor.max())
         tensor = (tensor - tensor.min()) / (tensor.max() - tensor.min())
+        tensor = torch.where(tensor > (threshold), tensor, torch.tensor(0, dtype=torch.float))
+        
 
         tensor = F.interpolate(tensor.unsqueeze(1), size=(H, W), mode='nearest')
         tensor = tensor.squeeze(1)
 
         self.model.to(offload_device)
-        results = tensor.cpu().float()
-        print(results.min(), results.max())
         
         if binary_mask:
             tensor = (tensor > 0).float()
         if blur_sigma > 0:
-            kernel_size = int(6 * blur_sigma + 1) 
+            kernel_size = int(6 * int(blur_sigma) + 1) 
             blur = transforms.GaussianBlur(kernel_size=(kernel_size, kernel_size), sigma=(blur_sigma, blur_sigma))
             tensor = blur(tensor)
 
@@ -105,8 +120,58 @@ Segments an image or batch of images using CLIPSeg.
         del outputs
         model_management.soft_empty_cache()
 
+        if prev_mask is not None:
+            tensor = tensor + prev_mask
+            torch.clamp(tensor, min=0.0, max=1.0)
+
         return tensor,
 
+class DownloadAndLoadCLIPSeg:
+
+    def __init__(self):
+        pass
+    
+    @classmethod
+    def INPUT_TYPES(s):
+       
+        return {"required":
+                    {     
+                    "model": (
+                    [   'Kijai/clipseg-rd64-refined-fp16',
+                        'CIDAS/clipseg-rd64-refined',
+                    ],
+                    {
+                    "default": 'clipseg-rd64-refined-fp16'
+                    }),
+                     },
+                }
+
+    CATEGORY = "KJNodes/masking"
+    RETURN_TYPES = ("CLIPSEGMODEL",)
+    RETURN_NAMES = ("clipseg_model",)
+    FUNCTION = "segment_image"
+    DESCRIPTION = """
+Downloads and loads CLIPSeg model with huggingface_hub,  
+to ComfyUI/models/clip_seg
+"""
+
+    def segment_image(self, model):
+        from transformers import CLIPSegProcessor, CLIPSegForImageSegmentation
+        checkpoint_path = os.path.join(folder_paths.models_dir,'clip_seg', model)
+        if not hasattr(self, "model"):
+            if not os.path.exists(checkpoint_path):
+                from huggingface_hub import snapshot_download
+                snapshot_download(repo_id=model, local_dir=checkpoint_path.split("/")[-1], local_dir_use_symlinks=False)
+            self.model = CLIPSegForImageSegmentation.from_pretrained(checkpoint_path)
+
+        processor = CLIPSegProcessor.from_pretrained(checkpoint_path)
+
+        clipseg_model = {}
+        clipseg_model['model'] = self.model
+        clipseg_model['processor'] = processor
+
+        return clipseg_model,
+
 class CreateTextMask:
 
     RETURN_TYPES = ("IMAGE", "MASK",)

requirements.txt

@@ -4,3 +4,4 @@ pillow>=10.3.0
 scipy
 color-matcher
 matplotlib
+huggingface_hub