HuPER: A Human-Inspired Framework for Phonetic Perception

arXiv:2602.01634 · eess.AS, cs.AI · Submitted 2026-02-02 · Read on arXiv

eess.AS, cs.AI

Submitted: 2026-02-02

Updated: 2026-09-25

Comments: NeurIPS 2026

Code: https://github.com/HuPER29/HuPER

License: http://creativecommons.org/licenses/by/4.0/

The gist: We propose HuPER, a human-inspired framework that models phonetic perception as adaptive inference over acoustic-phonetics evidence and linguistic knowledge.

Terminology

Abstract

We propose HuPER, a human-inspired framework that models phonetic perception as adaptive inference over acoustic-phonetics evidence and linguistic knowledge. With only 100 hours of training data, HuPER achieves state-of-the-art phonetic error rates on five English benchmarks and strong zero-shot transfer to 95 unseen languages. HuPER is also the first framework to enable adaptive, multi-path phonetic perception under diverse acoustic conditions. All training data, models, and code are open-sourced. Code and demo avaliable at https://github.com/Berkeley-Speech-Group/HuPER.

Sources

Related papers