Update

app.py

@@ -20,6 +20,8 @@ DESCRIPTION = '''# Text2Human
 
 This is an unofficial demo for <a href="https://github.com/yumingj/Text2Human">https://github.com/yumingj/Text2Human</a>.
 You can modify sample steps and seeds. By varying seeds, you can sample different human images under the same pose, shape description, and texture description. The larger the sample steps, the better quality of the generated images. (The default value of sample steps is 256 in the original repo.)
+
+Label image generation step can be skipped. However, in that case, the input label image must be 512x256 in size and must contain only the specified colors.
 '''
 FOOTER = '<img id="visitor-badge" alt="visitor badge" src="https://visitor-badge.glitch.me/badge?page_id=hysts.text2human" />'
 
@@ -68,8 +70,7 @@ def main():
                 with gr.Row():
                     label_image = gr.Image(label='Label Image',
                                            type='numpy',
-                                           elem_id='label-image',
-                                           interactive=False)
+                                           elem_id='label-image')
                 with gr.Row():
                     shape_text = gr.Textbox(
                         label='Shape Description',

model.py

@@ -52,6 +52,7 @@ class Model:
         self.config['device'] = device
         self._download_models()
         self.model = SampleFromPoseModel(self.config)
+        self.model.batch_size = 1
 
     def _load_config(self) -> dict:
         path = 'Text2Human/configs/sample_from_pose.yml'
@@ -84,10 +85,13 @@ class Model:
 
     @staticmethod
     def process_mask(mask: torch.Tensor) -> torch.Tensor:
+        if mask.shape != (512, 256, 3):
+            return None
         seg_map = np.full(mask.shape[:-1], -1)
         for index, color in enumerate(COLOR_LIST):
             seg_map[np.sum(mask == color, axis=2) == 3] = index
-        assert (seg_map != -1).all()
+        if not (seg_map != -1).all():
+            return None
         return seg_map
 
     @staticmethod
@@ -124,6 +128,8 @@ class Model:
             return
         mask = label_image.copy()
         seg_map = self.process_mask(mask)
+        if seg_map is None:
+            return
         self.model.segm = torch.from_numpy(seg_map).unsqueeze(0).unsqueeze(
             0).to(self.model.device)
         self.model.generate_quantized_segm()