Toward Streaming ASR with Non-Autoregressive Insertion-based Model

TL;DR

This paper introduces a unified neural network architecture combining audio segmentation and non-autoregressive insertion-based ASR to achieve high accuracy with low real-time factor in streaming speech recognition.

Contribution

It proposes a novel integrated model that performs audio segmentation and non-autoregressive ASR simultaneously, improving efficiency and accuracy over traditional separate models.

Findings

Achieved a better trade-off between accuracy and RTF compared to autoregressive baselines.

Demonstrated effectiveness on Japanese and English datasets.

Unified model reduces latency and complexity in streaming ASR systems.

Abstract

Neural end-to-end (E2E) models have become a promising technique to realize practical automatic speech recognition (ASR) systems. When realizing such a system, one important issue is the segmentation of audio to deal with streaming input or long recording. After audio segmentation, the ASR model with a small real-time factor (RTF) is preferable because the latency of the system can be faster. Recently, E2E ASR based on non-autoregressive models becomes a promising approach since it can decode an $N$-length token sequence with less than $N$ iterations. We propose a system to concatenate audio segmentation and non-autoregressive ASR to realize high accuracy and low RTF ASR. As a non-autoregressive ASR, the insertion-based model is used. In addition, instead of concatenating separated models for segmentation and ASR, we introduce a new architecture that realizes audio segmentation and non-autoregressive ASR by a single neural network. Experimental results on Japanese and English dataset show that the method achieved a reasonable trade-off between accuracy and RTF compared with baseline autoregressive Transformer and connectionist temporal classification.