DAME: Duration-Aware Matryoshka Embedding for Duration-Robust Speaker Verification
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917211776483328 |
|---|---|
| author | Jung, Youngmoon Yang, Joon-Young Kim, Ju-ho Roh, Jaeyoung Han, Chang Woo Cho, Hoon-Young |
| author_facet | Jung, Youngmoon Yang, Joon-Young Kim, Ju-ho Roh, Jaeyoung Han, Chang Woo Cho, Hoon-Young |
| contents | Short-utterance speaker verification remains challenging due to limited speaker-discriminative cues in short speech segments. While existing methods focus on enhancing speaker encoders, the embedding learning strategy still forces a single fixed-dimensional representation reused for utterances of any length, leaving capacity misaligned with the information available at different durations. We propose Duration-Aware Matryoshka Embedding (DAME), a model-agnostic framework that builds a nested hierarchy of sub-embeddings aligned to utterance durations: lower-dimensional representations capture compact speaker traits from short utterances, while higher dimensions encode richer details from longer speech. DAME supports both training from scratch and fine-tuning, and serves as a direct alternative to conventional large-margin fine-tuning, consistently improving performance across durations. On the VoxCeleb1-O/E/H and VOiCES evaluation sets, DAME consistently reduces the equal error rate on 1-s and other short-duration trials, while maintaining full-length performance with no additional inference cost. These gains generalize across various speaker encoder architectures under both general training and fine-tuning setups. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_13999 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DAME: Duration-Aware Matryoshka Embedding for Duration-Robust Speaker Verification Jung, Youngmoon Yang, Joon-Young Kim, Ju-ho Roh, Jaeyoung Han, Chang Woo Cho, Hoon-Young Audio and Speech Processing Artificial Intelligence Short-utterance speaker verification remains challenging due to limited speaker-discriminative cues in short speech segments. While existing methods focus on enhancing speaker encoders, the embedding learning strategy still forces a single fixed-dimensional representation reused for utterances of any length, leaving capacity misaligned with the information available at different durations. We propose Duration-Aware Matryoshka Embedding (DAME), a model-agnostic framework that builds a nested hierarchy of sub-embeddings aligned to utterance durations: lower-dimensional representations capture compact speaker traits from short utterances, while higher dimensions encode richer details from longer speech. DAME supports both training from scratch and fine-tuning, and serves as a direct alternative to conventional large-margin fine-tuning, consistently improving performance across durations. On the VoxCeleb1-O/E/H and VOiCES evaluation sets, DAME consistently reduces the equal error rate on 1-s and other short-duration trials, while maintaining full-length performance with no additional inference cost. These gains generalize across various speaker encoder architectures under both general training and fine-tuning setups. |
| title | DAME: Duration-Aware Matryoshka Embedding for Duration-Robust Speaker Verification |
| topic | Audio and Speech Processing Artificial Intelligence |
| url | https://arxiv.org/abs/2601.13999 |