Abstract
Serialized Output Training (SOT) has showcased state-of-the-art performance in multi-talker speech recognition by sequentially decoding the speech of individual speakers. To address the challenging label-permutation issue, prior methods have relied on either the Permutation Invariant Training (PIT) or the time-based First-In-First-Out (FIFO) rule. This study presents a model-based serialization strategy that incorporates an auxiliary module into the Attention Encoder-Decoder architecture, autonomously identifying the crucial factors to order the output sequence of the speech components in multi-talker speech. Experiments conducted on the LibriSpeech and LibriMix databases reveal that our approach significantly outperforms the PIT and FIFO baselines in both 2-mix and 3-mix scenarios. Further analysis shows that the serialization module identifies dominant speech components in a mixture by factors including loudness and gender, and orders speech components based on the dominance score.
| Original language | English |
|---|---|
| Pages (from-to) | 712-716 |
| Number of pages | 5 |
| Journal | Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH |
| DOIs | |
| State | Published - 2024 |
| Externally published | Yes |
| Event | 25th Interspeech Conferece 2024 - Kos Island, Greece Duration: 1 Sep 2024 → 5 Sep 2024 |
Keywords
- dominant speech
- multi-talker speech recognition
- serialized output training
Fingerprint
Dive into the research topics of 'Serialized Output Training by Learned Dominance'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver