update app.py

app.py

@@ -62,6 +62,11 @@ from  utils import computer_info
 from ram_utils import iou, sort_and_deduplicate, relation_classes, MLP, show_anns, ram_show_mask
 from ram_train_eval import RamModel,RamPredictor
 from mmengine.config import Config as mmengine_Config
+from lama_cleaner.helper import (
+    load_img,
+    numpy_to_bytes,
+    resize_max_size,
+)
 
 config_file = 'GroundingDINO/groundingdino/config/GroundingDINO_SwinT_OGC.py'
 ckpt_repo_id = "ShilongLiu/GroundingDINO"
@@ -126,14 +131,19 @@ def plot_boxes_to_image(image_pil, tgt):
             bbox = (x0, y0, w + x0, y0 + h)
         # bbox = draw.textbbox((x0, y0), str(label))
         draw.rectangle(bbox, fill=color)
-        font = os.path.join(cv2.__path__[0],'qt','fonts','DejaVuSans.ttf')
-        font_size = 36
-        new_font = ImageFont.truetype(font, font_size)
 
-        draw.text((x0+2, y0+2), str(label), font=new_font, fill="white")
+        try:
+            font = os.path.join(cv2.__path__[0],'qt','fonts','DejaVuSans.ttf')
+            font_size = 36
+            new_font = ImageFont.truetype(font, font_size)
+
+            draw.text((x0+2, y0+2), str(label), font=new_font, fill="white")
+        except Exception as e:
+            pass
 
         mask_draw.rectangle([x0, y0, x1, y1], fill=255, width=6)
 
+
     return image_pil, mask
 
 def load_image(image_path):
@@ -282,11 +292,6 @@ def load_lama_cleaner_model():
     # initialize lama_cleaner
     global lama_cleaner_model
     logger.info(f"initialize lama_cleaner...")
-    from lama_cleaner.helper import (
-        load_img,
-        numpy_to_bytes,
-        resize_max_size,
-    )
 
     lama_cleaner_model = ModelManager(
             name='lama',
@@ -522,16 +527,25 @@ def run_anything_task(input_image, text_prompt, task_type, inpaint_prompt, box_t
     file_temp = int(time.time())
     logger.info(f'run_anything_task_[{file_temp}]_{task_type}/{inpaint_mode}/[{mask_source_radio}]/{remove_mode}/{remove_mask_extend}_[{text_prompt}]/[{inpaint_prompt}]___1_')
 
+    output_images = []
+
     # load image
-    input_mask_pil = input_image['mask']
-    input_mask = np.array(input_mask_pil.convert("L"))  
-     
-    image_pil, image = load_image(input_image['image'].convert("RGB"))
+    if mask_source_radio == mask_source_draw:
+        input_mask_pil = input_image['mask']
+        input_mask = np.array(input_mask_pil.convert("L"))  
+    
+    print(type(input_image))
+    if isinstance(input_image, list):
+        image_pil, image = load_image(input_image['image'].convert("RGB"))
+        input_img = input_image['image']
+        output_images.append(input_image['image'])
+    else:
+        image_pil, image = load_image(input_image.convert("RGB"))
+        input_img = input_image
+        output_images.append(input_image)
 
     size = image_pil.size
-
-    output_images = []
-    output_images.append(input_image['image'])
+    
     # run grounding dino model
     if (task_type == 'inpainting' or task_type == 'remove') and mask_source_radio == mask_source_draw:
         pass
@@ -557,12 +571,13 @@ def run_anything_task(input_image, text_prompt, task_type, inpaint_prompt, box_t
             "size": [size[1], size[0]],  # H,W
             "labels": pred_phrases,
         }
+
         image_with_box = plot_boxes_to_image(copy.deepcopy(image_pil), pred_dict)[0]
         output_images.append(image_with_box)
 
     logger.info(f'run_anything_task_[{file_temp}]_{task_type}_2_')
     if task_type == 'segment' or ((task_type == 'inpainting' or task_type == 'remove') and mask_source_radio == mask_source_segment):
-        image = np.array(input_image['image'])
+        image = np.array(input_img)
         sam_predictor.set_image(image)
 
         H, W = size[1], size[0]