update: support scannet

open_vocab_seg/utils/predictor.py

@@ -762,8 +762,8 @@ class VisualizationDemoIndoor(VisualizationDemo):
                 frame = (255. * result.detach().cpu().squeeze().permute(1, 2, 0).numpy()).astype(np.uint8)
                 frames.append(frame)
 
-            video_out_file = '{}.gif'.format(name)
-            imageio.mimwrite(os.path.join('outputs', video_out_file), frames, fps=25)
+            # video_out_file = '{}.gif'.format(name)
+            # imageio.mimwrite(os.path.join('outputs', video_out_file), frames, fps=25)
             
             video_out_file = '{}.mp4'.format(name)
             imageio.mimwrite(os.path.join('outputs', video_out_file), frames, fps=25, quality=8)