SHNU Multilingual Conversational Speech Recognition System for INTERSPEECH 2025 MLC-SLM Challenge
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915376161357824 |
|---|---|
| author | Mei, Yuxiang Zheng, Yuang Xu, Dongxing Long, Yanhua |
| author_facet | Mei, Yuxiang Zheng, Yuang Xu, Dongxing Long, Yanhua |
| contents | This paper describes SHNU multilingual conversational speech recognition system (SHNU-mASR, team name-"maybe"), submitted to Track 1 of the INTERSPEECH 2025 MLC-SLM Challenge. Our system integrates a parallel-speech-encoder architecture with a large language model (LLM) to form a unified multilingual ASR framework. The parallel-speech-encoder consists of two pre-trained encoders, the Whisper-large-v3 encoder and mHuBERT-147 encoder. Their output embeddings are concatenated and fed into the LLM, enabling the model to leverage complementary acoustic and linguistic knowledge and achieve competitive performance. Moreover, we adopt a tri-stage training strategy to jointly update the low-rank adaptation modules and projector parameters of both the speech encoders and the LLM. In addition, we incorporate an additional language-aware prompt at the LLM input to enhance language-specific text generation. The SHNU-mASR system achieves an overall character/word error rate (CER/WER) of 11.76% on the blind evaluation set of the challenge, outperforming the official MLC-SLM baseline by 8.41 absolute CER/WER, without increasing the baseline training data. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_03343 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | SHNU Multilingual Conversational Speech Recognition System for INTERSPEECH 2025 MLC-SLM Challenge Mei, Yuxiang Zheng, Yuang Xu, Dongxing Long, Yanhua Computation and Language Audio and Speech Processing This paper describes SHNU multilingual conversational speech recognition system (SHNU-mASR, team name-"maybe"), submitted to Track 1 of the INTERSPEECH 2025 MLC-SLM Challenge. Our system integrates a parallel-speech-encoder architecture with a large language model (LLM) to form a unified multilingual ASR framework. The parallel-speech-encoder consists of two pre-trained encoders, the Whisper-large-v3 encoder and mHuBERT-147 encoder. Their output embeddings are concatenated and fed into the LLM, enabling the model to leverage complementary acoustic and linguistic knowledge and achieve competitive performance. Moreover, we adopt a tri-stage training strategy to jointly update the low-rank adaptation modules and projector parameters of both the speech encoders and the LLM. In addition, we incorporate an additional language-aware prompt at the LLM input to enhance language-specific text generation. The SHNU-mASR system achieves an overall character/word error rate (CER/WER) of 11.76% on the blind evaluation set of the challenge, outperforming the official MLC-SLM baseline by 8.41 absolute CER/WER, without increasing the baseline training data. |
| title | SHNU Multilingual Conversational Speech Recognition System for INTERSPEECH 2025 MLC-SLM Challenge |
| topic | Computation and Language Audio and Speech Processing |
| url | https://arxiv.org/abs/2507.03343 |