TY - JOUR
T1 - Hypernetworks for Personalizing ASR to Atypical Speech
AU - Müller-Eberstein, Max
AU - Yee, Dianna
AU - Yang, Karren
AU - Mantena, Gautam Varma
AU - Lea, Colin
PY - 2024/9
Y1 - 2024/9
N2 - Parameter-efficient fine-tuning (PEFT) for personalizing automatic speech recognition (ASR) has recently shown promise for adapting general population models to atypical speech. However, these approaches assume a priori knowledge of the atypical speech disorder being adapted for—the diagnosis of which requires expert knowledge that is not always available. Even given this knowledge, data scarcity and high inter-/intra-speaker variability further limit the effectiveness of traditional fine-tuning. To circumvent these challenges, we first identify the minimal set of model parameters required for ASR adaptation. Our analysis of each individual parameter’s effect on adaptation performance allows us to reduce Word Error Rate (WER) by half while adapting 0.03% of all weights. Alleviating the need for cohort-specific models, we next propose the novel use of a meta-learned hypernetwork to generate highly individualized, utterance-level adaptations on-the-fly for a diverse set of atypical speech characteristics. Evaluating adaptation at the global, cohort, and individual-level, we show that hypernetworks generalize better to out-of-distribution speakers, while maintaining an overall relative WER reduction of 75.2% using 0.1% of the full parameter budget.
AB - Parameter-efficient fine-tuning (PEFT) for personalizing automatic speech recognition (ASR) has recently shown promise for adapting general population models to atypical speech. However, these approaches assume a priori knowledge of the atypical speech disorder being adapted for—the diagnosis of which requires expert knowledge that is not always available. Even given this knowledge, data scarcity and high inter-/intra-speaker variability further limit the effectiveness of traditional fine-tuning. To circumvent these challenges, we first identify the minimal set of model parameters required for ASR adaptation. Our analysis of each individual parameter’s effect on adaptation performance allows us to reduce Word Error Rate (WER) by half while adapting 0.03% of all weights. Alleviating the need for cohort-specific models, we next propose the novel use of a meta-learned hypernetwork to generate highly individualized, utterance-level adaptations on-the-fly for a diverse set of atypical speech characteristics. Evaluating adaptation at the global, cohort, and individual-level, we show that hypernetworks generalize better to out-of-distribution speakers, while maintaining an overall relative WER reduction of 75.2% using 0.1% of the full parameter budget.
KW - Natural Language Processing
KW - Machine Learning
KW - Accessibility
KW - Speech Recognition
KW - Natural Language Processing
KW - Machine Learning
KW - Accessibility
KW - Speech Recognition
UR - https://machinelearning.apple.com/research/hypernetworks-personalizing-asr
U2 - 10.1162/tacl_a_00696
DO - 10.1162/tacl_a_00696
M3 - Journal article
SN - 2307-387X
VL - 12
SP - 1182
JO - Transactions of the Association for Computational Linguistics
JF - Transactions of the Association for Computational Linguistics
ER -