Correlation of Fréchet Audio Distance With Human Perception of Environmental Audio Is Embedding Dependant
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914729310552064 |
|---|---|
| author | Tailleur, Modan Lee, Junwon Lagrange, Mathieu Choi, Keunwoo Heller, Laurie M. Imoto, Keisuke Okamoto, Yuki |
| author_facet | Tailleur, Modan Lee, Junwon Lagrange, Mathieu Choi, Keunwoo Heller, Laurie M. Imoto, Keisuke Okamoto, Yuki |
| contents | This paper explores whether considering alternative domain-specific embeddings to calculate the Fréchet Audio Distance (FAD) metric can help the FAD to correlate better with perceptual ratings of environmental sounds. We used embeddings from VGGish, PANNs, MS-CLAP, L-CLAP, and MERT, which are tailored for either music or environmental sound evaluation. The FAD scores were calculated for sounds from the DCASE 2023 Task 7 dataset. Using perceptual data from the same task, we find that PANNs-WGM-LogMel produces the best correlation between FAD scores and perceptual ratings of both audio quality and perceived fit with a Spearman correlation higher than 0.5. We also find that music-specific embeddings resulted in significantly lower results. Interestingly, VGGish, the embedding used for the original Fréchet calculation, yielded a correlation below 0.1. These results underscore the critical importance of the choice of embedding for the FAD metric design. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_17508 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Correlation of Fréchet Audio Distance With Human Perception of Environmental Audio Is Embedding Dependant Tailleur, Modan Lee, Junwon Lagrange, Mathieu Choi, Keunwoo Heller, Laurie M. Imoto, Keisuke Okamoto, Yuki Sound Audio and Speech Processing This paper explores whether considering alternative domain-specific embeddings to calculate the Fréchet Audio Distance (FAD) metric can help the FAD to correlate better with perceptual ratings of environmental sounds. We used embeddings from VGGish, PANNs, MS-CLAP, L-CLAP, and MERT, which are tailored for either music or environmental sound evaluation. The FAD scores were calculated for sounds from the DCASE 2023 Task 7 dataset. Using perceptual data from the same task, we find that PANNs-WGM-LogMel produces the best correlation between FAD scores and perceptual ratings of both audio quality and perceived fit with a Spearman correlation higher than 0.5. We also find that music-specific embeddings resulted in significantly lower results. Interestingly, VGGish, the embedding used for the original Fréchet calculation, yielded a correlation below 0.1. These results underscore the critical importance of the choice of embedding for the FAD metric design. |
| title | Correlation of Fréchet Audio Distance With Human Perception of Environmental Audio Is Embedding Dependant |
| topic | Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2403.17508 |