arXiv cs.AI· Albert Zeyer, Davyd Naveriani, Dorian Koch, Ralf Schl\"uter, Hermann Ney·· 12 小时前AI 评分38
用每个音素一个均值 log-mel 向量替代 TTS：语音识别的简单快速文本注入方法
One Mean Log-Mel Vector per Phoneme: Simple and Fast Text Injection for Speech Recognition
AI 导读
研究用查表替代冻结 TTS 模型做文本注入：每个音素对应一个均值 log-mel 向量（或 one-hot 编码），按对齐统计随机采样时长重复，词间插入静音，训练速度比神经 TTS 快 1.8 倍且词错误率相当。
正文
Abstract:Text injection trains speech recognition models on unpaired text alongside paired audio. We feed the speech encoder synthetic input derived from text so the whole model learns from it. Previously, we synthesized this input with a frozen text-to-speech (TTS) model. Here, a simple lookup table replaces the TTS model: a mean log-mel vector (or one-hot encoding) per phoneme, repeated for a random duration drawn from alignment statistics, with silence between words. Training is 1.8 times faster than with the neural TTS, achieving comparable word error rates. With an external language model, our system reaches 1.36 / 2.90 / 1.52 / 3.24 % WER on LibriSpeech {dev,test}-{clean,other}, to our knowledge state-of-the-art using only LibriSpeech paired data, text, and lexicon. We analyze components of synthetic input (durations, acoustic features), compare different HMM aligners, evaluate scaling on the 25 000 h Loquacious dataset, and show that full-encoder injection yields strong relative improvements compared to partial-encoder and decoder methods. The code of the whole pipeline behind every number in this paper is public: this http URL
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE) |
| Cite as: | arXiv:2610.34623 [cs.AI] |
| (or arXiv:2610.34623v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.34623 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Albert Zeyer [view email]
[v1]
Sun, 25 Oct 2026 22:43:34 UTC (59 KB)
来源：arXiv cs.AI · arxiv.org