Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Jiang, Shan, Qi, Jiawen, Huo, Chuanbing, Gao, Yingqiang, Chen, Qinyu
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2603.15261
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917347178053632
author Jiang, Shan
Qi, Jiawen
Huo, Chuanbing
Gao, Yingqiang
Chen, Qinyu
author_facet Jiang, Shan
Qi, Jiawen
Huo, Chuanbing
Gao, Yingqiang
Chen, Qinyu
contents Personalizing automatic speech recognition (ASR) systems for non-normative speech, such as dysarthric and aphasic speech, is challenging. While speaker-specific fine-tuning (SS-FT) is widely used, it is typically initialized directly from a generic pre-trained model. Whether speaker-independent adaptation provides a stronger initialization prior under such mismatch remains unclear. In this work, we propose a two-stage adaptation framework consisting of speaker-independent fine-tuning (SI-FT) on multi-speaker non-normative data followed by SS-FT, and evaluate it through a controlled comparison with direct SS-FT under identical per-speaker conditions. Experiments on AphasiaBank and UA-Speech with Whisper-Large-v3 and Qwen3-ASR, alongside evaluation on typical-speech datasets TED-LIUM v3 and FLEURS, show that two-stage adaptation consistently improves personalization while maintaining manageable out-of-domain (OOD) trade-offs.
format Preprint
id arxiv_https___arxiv_org_abs_2603_15261
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Two-Stage Adaptation for Non-Normative Speech Recognition: Revisiting Speaker-Independent Initialization for Personalization
Jiang, Shan
Qi, Jiawen
Huo, Chuanbing
Gao, Yingqiang
Chen, Qinyu
Sound
Personalizing automatic speech recognition (ASR) systems for non-normative speech, such as dysarthric and aphasic speech, is challenging. While speaker-specific fine-tuning (SS-FT) is widely used, it is typically initialized directly from a generic pre-trained model. Whether speaker-independent adaptation provides a stronger initialization prior under such mismatch remains unclear. In this work, we propose a two-stage adaptation framework consisting of speaker-independent fine-tuning (SI-FT) on multi-speaker non-normative data followed by SS-FT, and evaluate it through a controlled comparison with direct SS-FT under identical per-speaker conditions. Experiments on AphasiaBank and UA-Speech with Whisper-Large-v3 and Qwen3-ASR, alongside evaluation on typical-speech datasets TED-LIUM v3 and FLEURS, show that two-stage adaptation consistently improves personalization while maintaining manageable out-of-domain (OOD) trade-offs.
title Two-Stage Adaptation for Non-Normative Speech Recognition: Revisiting Speaker-Independent Initialization for Personalization
topic Sound
url https://arxiv.org/abs/2603.15261