Vision-to-Audio Retrieval on AudioSet Eval
39.5Recall@1VAB-Encodec
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| VAB-EncodecEvaluation Protocol=Fine-tuned2026.04 | 39.5 | 65.4 | 74.6 | |
| TG-DPEvaluation Protocol=Zero-shot2026.04 | 37.4 | 59.2 | 68.1 | |
| CAV-MAE SyncEvaluation Protocol=Zero-shot2026.04 | 35.2 | 58.3 | 67.6 | |
| DenseAVEvaluation Protocol=Zero-shot2026.04 | 26.3 | 49.9 | 58.5 | |
| ImageBindEvaluation Protocol=Zero-shot2026.04 | 22.1 | 43.2 | 52.6 | |
| AVSiamEvaluation Protocol=Zero-shot2026.04 | 19.7 | — | — | |
| CAV-MAE Scale+++Evaluation Protocol=Zero-shot2026.04 | 18.8 | 39.5 | 50.1 | |
| CAV-MAEEvaluation Protocol=Zero-shot2026.04 | 16.1 | 38.6 | 49.3 | |
| LanguageBindEvaluation Protocol=Zero-shot2026.04 | 6.4 | 20.2 | 28.3 |