MOSS Transcribe Diarize Technical Report
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866911383694606336 |
|---|---|
| author | AI, MOSI. : Yu, Donghua Lin, Zhengyuan Yang, Chen Zhang, Yiyang Chen, Hanfu Chen, Jingqi Chen, Ke Fan, Liwei Jiang, Yi Zhu, Jie Li, Muchen Wang, Wenxuan Wang, Yang Xu, Zhe Gong, Yitian Zhang, Yuqian Zhang, Wenbo Wang, Songlin Wu, Zhiyu Fei, Zhaoye Cheng, Qinyuan Li, Shimin Qiu, Xipeng |
| author_facet | AI, MOSI. : Yu, Donghua Lin, Zhengyuan Yang, Chen Zhang, Yiyang Chen, Hanfu Chen, Jingqi Chen, Ke Fan, Liwei Jiang, Yi Zhu, Jie Li, Muchen Wang, Wenxuan Wang, Yang Xu, Zhe Gong, Yitian Zhang, Yuqian Zhang, Wenbo Wang, Songlin Wu, Zhiyu Fei, Zhaoye Cheng, Qinyuan Li, Shimin Qiu, Xipeng |
| contents | Speaker-Attributed, Time-Stamped Transcription (SATS) aims to transcribe what is said and to precisely determine the timing of each speaker, which is particularly valuable for meeting transcription. Existing SATS systems rarely adopt an end-to-end formulation and are further constrained by limited context windows, weak long-range speaker memory, and the inability to output timestamps. To address these limitations, we present MOSS Transcribe Diarize, a unified multimodal large language model that jointly performs Speaker-Attributed, Time-Stamped Transcription in an end-to-end paradigm. Trained on extensive real wild data and equipped with a 128k context window for up to 90-minute inputs, MOSS Transcribe Diarize scales well and generalizes robustly. Across comprehensive evaluations, it outperforms state-of-the-art commercial systems on multiple public and in-house benchmarks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_01554 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MOSS Transcribe Diarize Technical Report AI, MOSI. : Yu, Donghua Lin, Zhengyuan Yang, Chen Zhang, Yiyang Chen, Hanfu Chen, Jingqi Chen, Ke Fan, Liwei Jiang, Yi Zhu, Jie Li, Muchen Wang, Wenxuan Wang, Yang Xu, Zhe Gong, Yitian Zhang, Yuqian Zhang, Wenbo Wang, Songlin Wu, Zhiyu Fei, Zhaoye Cheng, Qinyuan Li, Shimin Qiu, Xipeng Sound Artificial Intelligence Audio and Speech Processing Speaker-Attributed, Time-Stamped Transcription (SATS) aims to transcribe what is said and to precisely determine the timing of each speaker, which is particularly valuable for meeting transcription. Existing SATS systems rarely adopt an end-to-end formulation and are further constrained by limited context windows, weak long-range speaker memory, and the inability to output timestamps. To address these limitations, we present MOSS Transcribe Diarize, a unified multimodal large language model that jointly performs Speaker-Attributed, Time-Stamped Transcription in an end-to-end paradigm. Trained on extensive real wild data and equipped with a 128k context window for up to 90-minute inputs, MOSS Transcribe Diarize scales well and generalizes robustly. Across comprehensive evaluations, it outperforms state-of-the-art commercial systems on multiple public and in-house benchmarks. |
| title | MOSS Transcribe Diarize Technical Report |
| topic | Sound Artificial Intelligence Audio and Speech Processing |
| url | https://arxiv.org/abs/2601.01554 |