A Comparison Of Semi-supervised Learning Techniques For Streaming ASR At Scale

Abstract

Unpaired text and audio injection have emerged as dominant methods for improving ASR performance in the absence of a large labeled corpus. However, little guidance exists on deploying these methods to improve production ASR systems that are trained on very large supervised corpora and with realistic requirements like a constrained model size and CPU budget, streaming capability, and a rich lattice for rescoring and for downstream NLU tasks. In this work, we compare three state-of-the-art semi-supervised methods encompassing both unpaired text and audio as well as several of their combinations in a controlled setting using joint training. We find that in our setting these methods offer many improvements beyond raw WER, including substantial gains in tail-word WER, decoder computation during inference, and lattice density.

Stats

• citations1
• S2 citations—
• github stars0
• HF likes0
• heat score2.26
• arxiv keypeyser2023a

Related papers

• Improving Streaming Transformer Based ASR Under A Framework Of Self-supervised Learning (2021)8.09
• Semi-supervised Sequence-to-sequence ASR Using Unpaired Speech And Text (2019)0.00
• Large Scale Weakly And Semi-supervised Learning For Low-resource Video ASR (2020)0.00
• Token-level Serialized Output Training For Joint Streaming ASR And ST Leveraging Textual Alignments (2023)0.00
• Improving Streaming Automatic Speech Recognition With Non-streaming Model Distillation On Unsupervised Data (2020)0.00
• On Comparison Of Encoders For Attention Based End To End Speech Recognition In Standalone And Rescoring Mode (2022)2.26
• Benchmarking LF-MMI, CTC And RNN-T Criteria For Streaming ASR (2020)8.35
• Training ASR Models By Generation Of Contextual Information (2019)0.00