hmchuong / MaGGIe

[CVPR24] MaGGIe: Mask Guided Gradual Human Instance Matting
https://maggie-matt.github.io/
Other
42 stars 3 forks source link

Error: 'No humans in first frame' for provided examples #7

Open Kamlesh364 opened 1 week ago

Kamlesh364 commented 1 week ago

I'm running the demo app implemented in Gardio. I am using the given examples only as input files but it works only on images and not on videos.

image

Terminal errors are as following-

Traceback (most recent call last):
  File "/home/kamlesh/.local/lib/python3.8/site-packages/gradio/queueing.py", line 536, in process_events
    response = await route_utils.call_process_api(
  File "/home/kamlesh/.local/lib/python3.8/site-packages/gradio/route_utils.py", line 322, in call_process_api
    output = await app.get_blocks().process_api(
  File "/home/kamlesh/.local/lib/python3.8/site-packages/gradio/blocks.py", line 1935, in process_api
    result = await self.call_function(
  File "/home/kamlesh/.local/lib/python3.8/site-packages/gradio/blocks.py", line 1520, in call_function
    prediction = await anyio.to_thread.run_sync(  # type: ignore
  File "/home/kamlesh/.local/lib/python3.8/site-packages/anyio/to_thread.py", line 56, in run_sync
    return await get_async_backend().run_sync_in_worker_thread(
  File "/home/kamlesh/.local/lib/python3.8/site-packages/anyio/_backends/_asyncio.py", line 2364, in run_sync_in_worker_thread
    return await future
  File "/home/kamlesh/.local/lib/python3.8/site-packages/anyio/_backends/_asyncio.py", line 864, in run
    result = context.run(func, *args)
  File "/home/kamlesh/.local/lib/python3.8/site-packages/gradio/utils.py", line 826, in wrapper
    response = f(*args, **kwargs)
  File "/home/kamlesh/MaGGIe/demo/app.py", line 39, in inference_video
    raise gr.Error("No person detected, please try video with visible human in the first frame!")
gradio.exceptions.Error: 'No person detected, please try video with visible human in the first frame!'
hmchuong commented 6 days ago

Hi, the demo requires the first frame to have visible people. There may be some problems with the human segmentation model that cannot detect people. I cannot fix it right now due to a deadline coming up. I will fix it later.