Group-Aware Partial Model Merging for Children's Automatic Speech Recognition

TL;DR

This paper presents GRAPAM, a novel group-aware partial model merging technique that improves children's speech recognition by clustering data, partially fine-tuning models, and merging them, resulting in better accuracy with fewer parameters.

Contribution

It introduces a group-aware partial model merging method that enhances children's ASR by combining clustering, partial fine-tuning, and model merging, outperforming full fine-tuning.

Findings

Achieves 6% relative WER reduction on MyST corpus.

Outperforms full fine-tuning with fewer parameters.

Effective in capturing group-specific characteristics.

Abstract

While supervised fine-tuning of adult pre-trained models for children's ASR has shown promise, it often fails to capture group-specific characteristics and variations among children. To address this, we introduce GRoup-Aware PARtial model Merging, a parameter-efficient approach that combines unsupervised clustering, partial fine-tuning, and model merging. Our approach adapts adult-pre-trained models to children by first grouping the children's data based on acoustic similarity. Each group is used to partially fine-tune an adult pre-trained model, and the resulting models are merged at the parameter level. Experiments conducted on the MyST children's speech corpus indicate that GRAPAM achieves a relative WER improvement of 6%, using the same amount of data, outperforming full fine-tuning while training fewer parameters.