TEVR: Improving Speech Recognition by Token Entropy Variance Reduction

25 Jun 2022  ·  Hajo Nils Krabbenhöft, Erhardt Barth ·

This paper presents TEVR, a speech recognition model designed to minimize the variation in token entropy w.r.t. to the language model. This takes advantage of the fact that if the language model will reliably and accurately predict a token anyway, then the acoustic model doesn't need to be accurate in recognizing it. We train German ASR models with 900 million parameters and show that on CommonVoice German, TEVR scores a very competitive 3.64% word error rate, which outperforms the best reported results by a relative 16.89% reduction in word error rate. We hope that releasing our fully trained speech recognition pipeline to the community will lead to privacy-preserving offline virtual assistants in the future.

PDF Abstract

Results from the Paper


 Ranked #1 on Speech Recognition on Common Voice German (using extra training data)

     Get a GitHub badge
Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Result Benchmark
Speech Recognition Common Voice German wav2vec 2.0 XLS-R 1B + TEVR (5-gram) Test WER 3.64% # 1
Test CER 1.54% # 2
Speech Recognition Common Voice German wav2vec 2.0 XLS-R 1B + TEVR (4-gram) Test WER 3.70% # 2
Speech Recognition Common Voice German wav2vec 2.0 XLS-R 1B (5-gram) Test WER 4.38% # 4
Test CER 1.62% # 3
Speech Recognition Common Voice German wav2vec 2.0 XLS-R 1B + TEVR (no LM) Test WER 10.10% # 13
Speech Recognition Common Voice German wav2vec 2.0 XLS-R (no LM) Test WER 12.06% # 14

Methods


No methods listed for this paper. Add relevant methods here