Visual Acoustic Matching on LibriSpeech unseen environments (test)
0.21RTE (s)LeMARA
Evaluation results
| Method | Links | |
|---|---|---|
| LeMARATrain dataset=AVSpeech-Rooms, visual_features=true2023.07 | 0.21 | |
| LeMARA (no vis)Train dataset=AVSpeech-Rooms, visual_features=false2023.07 | 0.223 | |
| AVITARTrain dataset=AVSpeech-Rooms2023.07 | 0.239 | |
| ViGASTrain dataset=AVSpeech-Rooms2023.07 | 0.254 | |
| Input audioTrain dataset=AVSpeech-Rooms2023.07 | 0.358 |