Modern speech recognition systems exhibit human-like word recognition but superhuman use of statistical dependencies between words

Advances in machine learning have yielded machine systems that approach human performance on real-world perception tasks. For speech recognition, it remains unclear whether such models' behavior resembles that of humans. We addressed this question by measuring word recognition abilities in humans and models across 272 speech distortions, providing a detailed behavioral fingerprint. Three recent industry speech recognition models trained on hundreds of years of speech outperformed humans. However, in conditions where linguistic context was removed, they matched human performance well, suggesting that the models are superhuman in their ability to leverage statistical dependencies between words. Experiments manipulating model training data suggest that training data variability and realism contribute to the match to human perception. Overall, the results show that superhuman performance is now possible with systems that are optimized over enough data and suggest that accurate models of human speech recognition require optimization over appropriate tasks and amounts of speech.

Authors

Institutions

Publication Details

Journal
bioRxiv (Cold Spring Harbor Laboratory)
Published
2026-11-18
DOI
https://doi.org/10.64898/2026.11.17.767420
Primary Topic
Speech Recognition and Synthesis
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Modern speech recognition systems exhibit human-like word recognition but superhuman use of statistical dependencies between words

Annika Magaro, Mark R. Saddler, Gasser Elbanna, Erica Shook et al.
2 links
bioRxiv (Cold Spring Harbor Laboratory)
Speech Recognition and Synthesis
preprint

Modern speech recognition systems exhibit human-like word recognition but superhuman use of statistical dependencies between words

Annika Magaro, Mark R. Saddler, Gasser Elbanna, Erica Shook, Josh H. McDermott, Alex Kell
preprint en

Abstract

Advances in machine learning have yielded machine systems that approach human performance on real-world perception tasks. For speech recognition, it remains unclear whether such models' behavior resembles that of humans. We addressed this question by measuring word recognition abilities in humans and models across 272 speech distortions, providing a detailed behavioral fingerprint. Three recent industry speech recognition models trained on hundreds of years of speech outperformed humans. However, in conditions where linguistic context was removed, they matched human performance well, suggesting that the models are superhuman in their ability to leverage statistical dependencies between words. Experiments manipulating model training data suggest that training data variability and realism contribute to the match to human perception. Overall, the results show that superhuman performance is now possible with systems that are optimized over enough data and suggest that accurate models of human speech recognition require optimization over appropriate tasks and amounts of speech.

bioRxiv (Cold Spring Harbor Laboratory)
Massachusetts Institute of Technology (US), Columbia University (US), Technical University of Denmark (DK)
Speech Recognition and Synthesis
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon