Controllable Time-Delay Transformer for Real-Time Punctuation Prediction and Disfluency Detection

TL;DR

This paper introduces a Controllable Time-delay Transformer that performs real-time punctuation prediction and disfluency detection, improving readability and downstream task performance in speech transcripts with minimal latency.

Contribution

The paper presents a novel CT-Transformer model with controllable delay and a fast decoding strategy for real-time joint punctuation and disfluency tasks, outperforming previous models.

Findings

Outperforms state-of-the-art on F-scores

Achieves competitive inference speed

Effective in real-time applications

Abstract

With the increased applications of automatic speech recognition (ASR) in recent years, it is essential to automatically insert punctuation marks and remove disfluencies in transcripts, to improve the readability of the transcripts as well as the performance of subsequent applications, such as machine translation, dialogue systems, and so forth. In this paper, we propose a Controllable Time-delay Transformer (CT-Transformer) model that jointly completes the punctuation prediction and disfluency detection tasks in real time. The CT-Transformer model facilitates freezing partial outputs with controllable time delay to fulfill the real-time constraints in partial decoding required by subsequent applications. We further propose a fast decoding strategy to minimize latency while maintaining competitive performance. Experimental results on the IWSLT2011 benchmark dataset and an in-house Chinese annotated dataset demonstrate that the proposed approach outperforms the previous state-of-the-art models on F-scores and achieves a competitive inference speed.