JimWest / MeFaMo

MIT License
480 stars 103 forks source link

Using --input video file #18

Open BTakacs-Oxit opened 1 year ago

Qaanaaq commented 1 year ago

Hello, unfortunately I also was not able to use an input video file for some reason. Was also trying the --input way but did not seem to work.

redbean commented 1 year ago

in my case I change the line 133 in mefamo.py like below cap = cv2.VideoCapture(input)#, cv2.CAP_DSHOW)

when the video is given in path, cv2.videocapture with cv2.CAP_DSHOW returns None.

nmlen commented 9 months ago

in my case I change the line 133 in mefamo.py like below cap = cv2.VideoCapture(input)#, cv2.CAP_DSHOW)

when the video is given in path, cv2.videocapture with cv2.CAP_DSHOW returns None.

Tried this, but getting a slightly different error at the exact same line.

image

Starting MeFaMo INFO: Created TensorFlow Lite XNNPACK delegate for CPU. Traceback (most recent call last): File "C:\Users\Main\MeFaMo\examples\mefamo_cli.py", line 22, in mediapipe_face.start() File "C:\Users\Main\MeFaMo\venv\lib\site-packages\mefamo_mediapipefacemocap-0.1-py3.10.egg\mefamo\mefamo.py", line 133, in start cap = cv2.VideoCapture(input)#, cv2.CAP_DSHOW) UnboundLocalError: local variable 'input' referenced before assignment