Segment-level Vectorized Beam Search Based On Partially Autoregressive Inference

Abstract

Attention-based encoder-decoder models with autoregressive (AR) decoding have proven to be the dominant approach for automatic speech recognition (ASR) due to their superior accuracy. However, they often suffer from slow inference. This is primarily attributed to the incremental calculation of the decoder. This work proposes a partially AR framework, which employs segment-level vectorized beam search for improving the inference speed of an ASR model based on the hybrid connectionist temporal classification (CTC) attention-based architecture. It first generates an initial hypothesis using greedy CTC decoding, identifying low-confidence tokens based on their output probabilities. We then utilize the decoder to perform segment-level vectorized beam search on these tokens, re-predicting in parallel with minimal decoder calculations. Experimental results show that our method is 12 to 13 times faster in inference on the LibriSpeech corpus over AR decoding whilst preserving high accuracy.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keysomeki2023segment

Related papers

• Vectorization Of Hypotheses And Speech For Faster Beam Search In Encoder Decoder-based Speech Recognition (2018)0.00
• Robust Beam Search For Encoder-decoder Attention Based Speech Recognition Without Length Bias (2020)4.52
• Streaming Parallel Transducer Beam Search With Fast-slow Cascaded Encoders (2022)0.00
• Joint Beam Search Integrating CTC, Attention, And Transducer Decoders (2024)5.24
• Integration Of Frame- And Label-synchronous Beam Search For Streaming Encoder-decoder Speech Recognition (2023)0.00
• Run-and-back Stitch Search: Novel Block Synchronous Decoding For Streaming Encoder-decoder ASR (2022)3.58
• A Fully Differentiable Beam Search Decoder (2019)0.00
• Towards Effective And Efficient Non-autoregressive Decoding Using Block-based Attention Mask (2024)0.00