# Active and Semi-Supervised Learning in ASR: Benefits on the Acoustic and Language Models

## TL;DR

This paper demonstrates that combining active learning with semi-supervised training significantly reduces transcription costs and improves speech recognition accuracy by optimizing data selection for acoustic and language models.

## Contribution

It introduces a confidence filtering-based data selection method that enhances both acoustic and language models in speech recognition, showing benefits over random selection.

## Key findings

- Active learning reduces transcription costs by about 70%.
- Combining AL and SST improves word error rate by approximately 12.5%.
- AL benefits both acoustic and language model training.

## Abstract

The goal of this paper is to simulate the benefits of jointly applying active learning (AL) and semi-supervised training (SST) in a new speech recognition application. Our data selection approach relies on confidence filtering, and its impact on both the acoustic and language models (AM and LM) is studied. While AL is known to be beneficial to AM training, we show that it also carries out substantial improvements to the LM when combined with SST. Sophisticated confidence models, on the other hand, did not prove to yield any data selection gain. Our results indicate that, while SST is crucial at the beginning of the labeling process, its gains degrade rapidly as AL is set in place. The final simulation reports that AL allows a transcription cost reduction of about 70% over random selection. Alternatively, for a fixed transcription budget, the proposed approach improves the word error rate by about 12.5% relative.