Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.18220 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917233165336576 |
|---|---|
| author | Mu, Bingshen Shi, Xian Wang, Xiong Liu, Hexin Xu, Jin Xie, Lei |
| author_facet | Mu, Bingshen Shi, Xian Wang, Xiong Liu, Hexin Xu, Jin Xie, Lei |
| contents | Forced alignment (FA) predicts start and end timestamps for words or characters in speech, but existing methods are language-specific and prone to cumulative temporal shifts. The multilingual speech understanding and long-sequence processing abilities of speech large language models (SLLMs) make them promising for FA in multilingual, crosslingual, and long-form speech settings. However, directly applying the next-token prediction paradigm of SLLMs to FA results in hallucinations and slow inference. To bridge the gap, we propose LLM-ForcedAligner, reformulating FA as a slot-filling paradigm: timestamps are treated as discrete indices, and special timestamp tokens are inserted as slots into the transcript. Conditioned on the speech embeddings and the transcript with slots, the SLLM directly predicts the time indices at slots. During training, causal attention masking with non-shifted input and label sequences allows each slot to predict its own timestamp index based on itself and preceding context, with loss computed only at slot positions. Dynamic slot insertion enables FA at arbitrary positions. Moreover, non-autoregressive inference is supported, avoiding hallucinations and improving speed. Experiments across multilingual, crosslingual, and long-form speech scenarios show that LLM-ForcedAligner achieves a 69%~78% relative reduction in accumulated averaging shift compared with prior methods. Checkpoint and inference code are available at https://github.com/QwenLM/Qwen3-ASR. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_18220 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | LLM-ForcedAligner: A Non-Autoregressive and Accurate LLM-Based Forced Aligner for Multilingual and Long-Form Speech Mu, Bingshen Shi, Xian Wang, Xiong Liu, Hexin Xu, Jin Xie, Lei Sound Audio and Speech Processing Forced alignment (FA) predicts start and end timestamps for words or characters in speech, but existing methods are language-specific and prone to cumulative temporal shifts. The multilingual speech understanding and long-sequence processing abilities of speech large language models (SLLMs) make them promising for FA in multilingual, crosslingual, and long-form speech settings. However, directly applying the next-token prediction paradigm of SLLMs to FA results in hallucinations and slow inference. To bridge the gap, we propose LLM-ForcedAligner, reformulating FA as a slot-filling paradigm: timestamps are treated as discrete indices, and special timestamp tokens are inserted as slots into the transcript. Conditioned on the speech embeddings and the transcript with slots, the SLLM directly predicts the time indices at slots. During training, causal attention masking with non-shifted input and label sequences allows each slot to predict its own timestamp index based on itself and preceding context, with loss computed only at slot positions. Dynamic slot insertion enables FA at arbitrary positions. Moreover, non-autoregressive inference is supported, avoiding hallucinations and improving speed. Experiments across multilingual, crosslingual, and long-form speech scenarios show that LLM-ForcedAligner achieves a 69%~78% relative reduction in accumulated averaging shift compared with prior methods. Checkpoint and inference code are available at https://github.com/QwenLM/Qwen3-ASR. |
| title | LLM-ForcedAligner: A Non-Autoregressive and Accurate LLM-Based Forced Aligner for Multilingual and Long-Form Speech |
| topic | Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2601.18220 |