arXiv：cs.CL· Minnie Kabra, Benjamin Lecouteux, Maximin Coavoux·· 3 小时前
网络何时可以剪枝？多语言语音解析中中间神经元的作用研究
When Can You Prune Your Network? A Study of Intermediate Neurons in Multilingual Speech Parsing
AI 导读
一项被 EMNLP 2026 Findings 收录的研究提出更简洁的端到端语音解析架构，移除中间神经网络单元后参数减少 12%，在 ASR 和解析任务上性能与先前方法相当或更优。研究发现，当预训练编码器被冻结时，中间 NN 单元有助于缩小表示差距；该研究在法语及中低资源语言斯洛文尼亚语和奈及利亚皮钦语上进行了全面评测，并考察了训练数据规模和预训练语音编码器中间层的影响。
正文
Abstract:End-to-end speech parsing, a task recently proposed, consists in predicting both the transcription and the syntactic tree for a spoken utterance. Existing architectures for speech parsing often utilise intermediate neural networks. In this work, we examine the effectiveness of intermediate neural networks (NN) for parsing, and, specifically, what role do they play. We introduce a simpler end-to-end architecture for speech parsing, where we remove these intermediate NN units, reducing the parameters by 12%, while achieving comparable or better performance than prior method on both automatic speech recognition (ASR) and parsing. We demonstrate that intermediate NN units help reduce the representational gap when the pre-trained encoder is frozen. We do a comprehensive evaluation of speech parsing on French, and medium-low resource languages Slovenian and Naija. We further investigate the impact of the training data size and intermediate layers of the pretrained speech encoder on speech parsing.
| Comments: | to appear in Findings of EMNLP 2026 |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.11520 [cs.CL] |
| (or arXiv:2610.11520v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.11520 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Maximin Coavoux [view email]
[v1]
Thu, 8 Oct 2026 08:53:41 UTC (356 KB)
来源：arXiv：cs.CL · arxiv.org