refactor audio-visual feature extraction

aiden200 / 2D3MF

Code and models for the paper "2D3MF: Deepfake Detection using Multi Modal Middle Fusion"

Other

31 stars 1 forks source link

Closed aromanusc closed 7 months ago

aromanusc commented 7 months ago

Refactor audio-visual feature extraction
Adds options for various audio feature extractors:
- Resnet (Done)
- xvector (Done)
- emotion2vec (TBD)
- MFCC (Done)

Requirements ⚠️:

The audio and video filenames should be the same, except for the .mp4 and .wav file extension.