research area Accessibility, research area Speech and Natural Language Processingconference Interspeech
content type paperpublished August 2025

The Interspeech 2025 Speech Accessibility Project Challenge

AuthorsXiuwen Zheng†, Bornali Phukon†, Jonghwan Na¶¶, Ed Cutrell‡, Kyu Jeong Han§, Mark Hasegawa-Johnson†, Pan-Pan Jiang¶, Aadrhik Khuila‡, Colin Lea, Bob MacDonald¶, Gautam Mantena, Venkatesh Ravichandran§, Leda Sari‡‡, Katrin Tomanek§§, Chang D. Yoo†††, Chris Zwilling†

View publication

While the last decade has witnessed significant advancements in Automatic Speech Recognition (ASR) systems, performance of these systems for individuals with speech disabilities remains inadequate, partly due to limited public training data. To bridge this gap, the 2025 Interspeech Speech Accessibility Project (SAP) Challenge was launched, utilizing over 400 hours of SAP data collected and transcribed from more than 500 individuals with diverse speech disabilities. Hosted on EvalAI and leveraging the remote evaluation pipeline, the SAP Challenge evaluates submissions based on Word Error Rate and Semantic Score. Consequently, 12 out of 22 valid teams outperformed the whisper-large-v2 baseline in terms of WER, while 17 teams surpassed the baseline on SemScore. Notably, the top team achieved the lowest WER of 8.11%, and the highest SemScore of 88.44% at the same time, setting new benchmarks for future ASR systems in recognizing impaired speech.

  • † University of Illinois Urbana-Champaign
  • ‡ Microsoft
  • § Amazon
  • ¶ Google
  • ‡‡ Meta
  • §§ CDLI
  • ¶¶ Inha University
  • ††† KAIST

Related readings and updates.

Affect Models Have Weak Generalizability to Atypical Speech

August 1, 2025research area Accessibility, research area Fairnessconference International Conference on Affective Computing and Intelligent Interaction (ACII)

Speech and voice conditions can alter the acoustic properties of speech, which could impact the performance of paralinguistic models for affect for people with atypical speech. We evaluate publicly available models for recognizing categorical and dimensional affect from speech on a dataset of atypical speech, comparing results to datasets of typical speech. We investigate three dimensions of speech atypicality: intelligibility, which is related...

Read more

Voice Quality Dimensions as Interpretable Primitives for Speaking Style for Atypical Speech and Affect

June 5, 2025research area Accessibility, research area Speech and Natural Language Processingconference Interspeech

Perceptual voice quality dimensions describe key characteristics of atypical speech and other speech modulations. Here we develop and evaluate voice quality models for seven voice and speech dimensions (intelligibility, imprecise consonants, harsh voice, naturalness, monoloudness, monopitch, and breathiness). Probes were trained on the public Speech Accessibility (SAP) project dataset with 11,184 samples from 434 speakers, using embeddings from...

Read more