DKU系统直接从原始音频识别多说话人语音，无需说话人标签。
The DKU System for Multi-Speaker Automatic Speech Recognition in MLC-SLM Challenge
- 将说话人嵌入与语句边界信息融入Qwen2.5大模型，实现端到端识别。
- 在开发集和测试集上分别达到23.56%和18.08%的tcpWER，显著优于基线。
- 适合需要高精度多说话人语音识别的工业应用与研究场景。
我们提出DKU系统参与MLC-SLM挑战赛任务2，目标是在无说话人标签或时间边界的情况下，直接从原始音频中进行多说话人自动语音识别。该方法基于一种包含说话人嵌入与时间语句边界的去重感知框架，并将其集成至基于Qwen2.5的大语言模型中。为进一步提升多语言性能，我们在大模型解码器中对语言特定适配器和LoRA模块进行微调。最终，系统在MLC-SLM数据集的开发集和测试集上分别取得23.56%和18.08%的tcpWER，显著优于官方基线。
原文摘要 · Abstract (English)
We present the DKU system for Task 2 of the MLC-SLM Challenge, which aims to perform multi-speaker automatic speech recognition directly from raw audio without Oracle speaker labels or time boundaries. Our approach builds upon a diarization-aware framework integrating speaker embeddings and temporal utterance boundaries into a Qwen2.5-based large language model (LLM). Then, we enhance the system's multilingual performance by fine-tuning language-specific adapters and LoRA modules within the LLM decoder. Finally, our system achieves the tcpWER of 23.56\% and 18.08\% on the development and test sets of the MLC-SLM dataset, substantially outperforming the official baseline.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。