Understanding Frechet Speech Distance for Synthetic Speech Quality Evaluation
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914289821941760 |
|---|---|
| author | Kim, June-Woo Agarwal, Dhruv Cerina, Federica |
| author_facet | Kim, June-Woo Agarwal, Dhruv Cerina, Federica |
| contents | Objective evaluation of synthetic speech quality remains a critical challenge. Human listening tests are the gold standard, but costly and impractical at scale. Fréchet Distance has emerged as a promising alternative, yet its reliability depends heavily on the choice of embeddings and experimental settings. In this work, we comprehensively evaluate Fréchet Speech Distance (FSD) and its variant Speech Maximum Mean Discrepancy (SMMD) under varied embeddings and conditions. We further incorporate human listening evaluations alongside TTS intelligibility and synthetic-trained ASR WER to validate the perceptual relevance of these metrics. Our findings show that WavLM Base+ features yield the most stable alignment with human ratings. While FSD and SMMD cannot fully replace subjective evaluation, we show that they can serve as complementary, cost-efficient, and reproducible measures, particularly useful when large-scale or direct listening assessments are infeasible. Code is available at https://github.com/kaen2891/FrechetSpeechDistance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_21386 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Understanding Frechet Speech Distance for Synthetic Speech Quality Evaluation Kim, June-Woo Agarwal, Dhruv Cerina, Federica Sound Artificial Intelligence Objective evaluation of synthetic speech quality remains a critical challenge. Human listening tests are the gold standard, but costly and impractical at scale. Fréchet Distance has emerged as a promising alternative, yet its reliability depends heavily on the choice of embeddings and experimental settings. In this work, we comprehensively evaluate Fréchet Speech Distance (FSD) and its variant Speech Maximum Mean Discrepancy (SMMD) under varied embeddings and conditions. We further incorporate human listening evaluations alongside TTS intelligibility and synthetic-trained ASR WER to validate the perceptual relevance of these metrics. Our findings show that WavLM Base+ features yield the most stable alignment with human ratings. While FSD and SMMD cannot fully replace subjective evaluation, we show that they can serve as complementary, cost-efficient, and reproducible measures, particularly useful when large-scale or direct listening assessments are infeasible. Code is available at https://github.com/kaen2891/FrechetSpeechDistance. |
| title | Understanding Frechet Speech Distance for Synthetic Speech Quality Evaluation |
| topic | Sound Artificial Intelligence |
| url | https://arxiv.org/abs/2601.21386 |