将口语中文直接转为正式书面语，模型更轻便，无需额外处理。
FormalASR: End-to-End Spoken Chinese to Formal Text
- 用大模型重写+筛选构建双语数据集，训练端到端模型。
- 在两个数据集上相对字符错误率降低37.4%，生成文本更流畅。
- 0.6B/1.7B小模型即可部署，适合手机等设备实时使用。
自动语音识别系统通常以逐字转录为目标，保留口语中的不连贯、填充词和非正式结构，这在下游写作应用中并不理想。常见做法是采用两阶段ASR+大模型后编辑，但会增加延迟与内存开销，难以在设备端部署。本文提出FormalASR，两款紧凑的端到端模型（0.6B和1.7B），可直接将口语中文转为正式书面语。为此，我们基于大模型重写与质量过滤，构建了两个大规模语料库：WenetSpeech-Formal和Speechio-Formal。随后，对Qwen3-ASR在两种规模下进行监督微调。在WenetSpeech-Formal和Speechio-Formal上的实验表明，FormalASR相比原始转录基线，字符错误率（CER）最高降低37.4%，同时提升ROUGE-L与BERTScore。该模型无需部署时的后处理大模型，提供轻量级、可本地运行的口语转正式文本解决方案。
原文摘要 · Abstract (English)
Automatic speech recognition (ASR) systems are typically optimized for verbatim transcription, which preserves disfluencies, filler words, and informal spoken structures that are often unsuitable for downstream writing-oriented applications. A common workaround is a two-stage ASR+LLM pipeline for post-editing, but this design increases latency and memory cost and is difficult to deploy on-device. We present FormalASR, two compact end-to-end models (0.6B and 1.7B) that directly transcribe spoken Chinese into formal written text. To enable this setting, we build WenetSpeech-Formal and Speechio-Formal, two large-scale spoken-to-formal datasets constructed by LLM-based rewriting and quality filtering. We then fine-tune Qwen3-ASR at two scales (0.6B and 1.7B) with supervised fine-tuning. Experiments on WenetSpeech-Formal and Speechio-Formal show that FormalASR achieves up to 37.4% relative CER reduction over verbatim baselines, while also improving ROUGE-L and BERTScore. FormalASR requires no post-processing LLM at deployment time, providing a lightweight, on-device solution for spoken-to-formal transcription.
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。