Merge branch 'master' of https://github.com/gaomingqi/Track-Anything

app.py

@@ -42,7 +42,7 @@ xmem_checkpoint = download_checkpoint(xmem_checkpoint_url, folder, xmem_checkpoi
 
 # args, defined in track_anything.py
 args = parse_augment()
-# args.port = 12213
+args.port = 12213
 model = TrackingAnything(SAM_checkpoint, xmem_checkpoint, args)
 
 
@@ -107,6 +107,9 @@ def get_frames_from_video(video_input, play_state):
     key_frame_index = int(timestamp * fps)
     nearest_frame = frames[key_frame_index]
     frames = [frames[:key_frame_index], frames[key_frame_index:], nearest_frame]
+
+    # set image in sam when select the template frame
+    model.samcontroler.set_image(np.asarray(nearest_frame))
     return frames, nearest_frame
 
 def inference_all(template_frame, point_prompt, click_state, evt:gr.SelectData):
@@ -127,13 +130,10 @@ def inference_all(template_frame, point_prompt, click_state, evt:gr.SelectData):
     # default value
     # points = np.array([[evt.index[0],evt.index[1]]])
     # labels= np.array([1])
-    mask, logit, painted_image = model.inference_step(first_flag=True, 
-                                                      interact_flag=False, 
+    mask, logit, painted_image = model.first_frame_click( 
                                                       image=np.asarray(template_frame), 
-                                                      same_image_flag=False,
                                                       points=np.array(prompt["input_point"]), 
                                                       labels=np.array(prompt["input_label"]),
-                                                      logits=None, 
                                                       multimask=prompt["multimask_output"]
                                                       )
     return painted_image, click_state