Deep Speech: Scaling up end-to-end speech recognition

Code

• github.com/PaddlePaddle/PaddleSpeech
  Officialpaddle★ 12,564
• github.com/Picovoice/stt-benchmark
  none★ 687
• github.com/Picovoice/speech-to-text-benchmark
  none★ 687
• github.com/msalhab96/SpeeQ
  pytorch★ 51
• github.com/anssssss/Vietnamese-Speech-Recognition
  tf★ 0
• github.com/bjtommychen/Keras_DeepSpeech2_SpeechRecognition
  tf★ 0
• github.com/GeorgeFedoseev/DeepSpeech
  tf★ 0
• github.com/RezisEwig/unity_speech
  none★ 0
• github.com/YuBeomGon/DeepSpeech
  tf★ 0
• github.com/WalterJohnson0/DeepSpeech-KerasRebuild
  tf★ 0

Abstract

We present a state-of-the-art speech recognition system developed using end-to-end deep learning. Our architecture is significantly simpler than traditional speech systems, which rely on laboriously engineered processing pipelines; these traditional systems also tend to perform poorly when used in noisy environments. In contrast, our system does not need hand-designed components to model background noise, reverberation, or speaker variation, but instead directly learns a function that is robust to such effects. We do not need a phoneme dictionary, nor even the concept of a "phoneme." Key to our approach is a well-optimized RNN training system that uses multiple GPUs, as well as a set of novel data synthesis techniques that allow us to efficiently obtain a large amount of varied data for training. Our system, called Deep Speech, outperforms previously published results on the widely studied Switchboard Hub5'00, achieving 16.0% error on the full test set. Deep Speech also handles challenging noisy environments better than widely used, state-of-the-art commercial speech systems.

Tasks

Benchmark Results

Reproductions