Image-to-Audio Retrieval on Flickr-SoundNet
4.92mAPC-MCR
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| C-MCRZero-shot=true, A-V Pairs=×, Trainable parameters=2.1M2023.05 | 4.92 | 1.58 | 5.98 | |
| AudioCLIPZero-shot=true, A-V Pairs=✓, Trainable parameters=134.1M2023.05 | 4.43 | 1.58 | 5.92 | |
| WAV2CLIPZero-shot=true, A-V Pairs=✓, Trainable parameters=11.7M2023.05 | 3.47 | 1.12 | 4.34 | |
| RandomZero-shot=true, A-V Pairs=-, Trainable parameters=-2023.05 | 0.17 | 0.02 | 0.06 |