search

andresprados / SPIGA

SPIGA: Shape Preserving Facial Landmarks with Graph Attention Networks.
https://bmvc2022.mpi-inf.mpg.de/155/
BSD 3-Clause "New" or "Revised" License
304 stars 32 forks source link

showing only the face with highest confidence #15

Open monajalal opened 1 year ago

monajalal commented 1 year ago

Hi Andrés Thanks a lot for the great repository. I have a mp4 video and some frames it shows two predicted face while there is only one face. When I am getting inference using your ipynb for video, is there a way I could force it to show only faces with probably of detection higher than 90%?

H1ppx commented 9 months ago

@monajalal I did some digging because I need something similar, the retinaface implementation does not seem to have any use of confidence at least in terms of the demo ./spiga/demo/analyze/track/retinasort/face_tracker.py. The import is from this repo so you can probably use some sort of modification to process_frame() in face_tracker.py to incorporate it

parsapahlavan commented 6 months ago

@monajalal This might be a bit late, but any other people facing the same question can try changing the 'config.py' file located at 'spiga/demo/analyze/track/retinasort'.

There are 'score_thr' attributes for the tracker you want to use.

emlcpfx commented 2 months ago

Is there a way to point it to the face you want? For example, if I used face recognition, how would I use that bbox or x,y coord to tell SPIGA which face I want?