Audio+Text2Audio+Text on Custom mixed modality dataset (test)
27.28R@1WavRAG
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| WavRAGAvg. Time=0.192025.02 | 27.28 | — | 51.84 | 63.13 | 43.81 | |
| CLAP (AT2AT)Avg. Time=0.092025.02 | 13.45 | — | 21.45 | 23.79 | 18.49 | |
| CLAP (AT)Avg. Time=0.052025.02 | 12.6 | — | 29.4 | 39.89 | 24.74 | |
| CLAP (TA)Avg. Time=0.052025.02 | 9.98 | — | 25.77 | 35.88 | 21.35 | |
| BGE (Caption)Avg. Time=1.992025.02 | 2.51 | — | 5.85 | 7.75 | 4.83 | |
| CLAP (ALL)Avg. Time=0.062025.02 | 0.01 | — | 0.12 | 0.18 | 0.02 |