Thank you and your team for the excellent work done for the Deepfake community; it's exactly what we needed. I extend my heartfelt congratulations on your achievements.
I've been testing my dataset with your project recently and wrote some additional code for preprocess.py to make it work with my data. However, I've encountered an issue: in many frames of certain videos, faces cannot be accurately detected. This might be due to the lower resolution of my video dataset. Additionally, I've noticed that a common characteristic of most failed cases is that the face occupies a large portion of the frame, with a relatively small background area.
Therefore, I'm seeking your advice on how to address this issue. Are there any parameters that can be adjusted to relax the conditions for face detection? Or, if you have any other suggestions, I'm also very open to them.
Hi Zhiyuan,
Thank you and your team for the excellent work done for the Deepfake community; it's exactly what we needed. I extend my heartfelt congratulations on your achievements.
I've been testing my dataset with your project recently and wrote some additional code for
preprocess.py
to make it work with my data. However, I've encountered an issue: in many frames of certain videos, faces cannot be accurately detected. This might be due to the lower resolution of my video dataset. Additionally, I've noticed that a common characteristic of most failed cases is that the face occupies a large portion of the frame, with a relatively small background area.Therefore, I'm seeking your advice on how to address this issue. Are there any parameters that can be adjusted to relax the conditions for face detection? Or, if you have any other suggestions, I'm also very open to them.
Best, Ruihan