Demographic Bias in ASR

ASR: Automatic Speech Recognition

Follow this topic Ask about this topic

Momentum

5 papers in the last four weeks, with none the four weeks before. 0.0% of all new papers.

Jul 13Week of Sep 28

Latest papers 16

All topics

InterCorrect: Intersection-Aware Correction of Demographic Model Merging for Fair ASR
Oct 6, 2026Ashley E. Bravo-Bravo, Yuchen Zhang, Haralambos Mouratidis +2Demographic Bias in ASR Model Merging
PUCP · IADS, University of Essex · CSEE, University of Essex +1
Pruning for Efficiency, Paying in Fairness: Demographic Disparities in Pruned Speech-LLMs
Sep 29, 2026Ganesh Pavan Kartikeya Bharadwaj Kolluri, Michael Kampouridis, Ravi ShekharLLM Pruning Algorithmic Fairness
School of Computer Science and Electronic Engineering, University of Essex, UK
Temporal Taxation Compounds Under Post-Training Compression of Whisper Models
Sep 23, 2026Srishti Ginjala, Eric Fosler-Lussier, Christopher W. Myers +1ASR Evaluation Speech Foundation Models
The Ohio State University, Columbus, OH, USA · Air Force Research Laboratory, USA
When Entanglement Lower-Bounds Disparity: Auditing and Repairing Demographic Fairness in Audio Understanding Models
Sep 23, 2026Kian Shamsaie, Iman ModarressiDisentangled Representation Learning Audio Representation Learning
People Make Things
Vimarsha: Faithful ASR Evaluation for Indian Languages with Demographic Diversity, In-the-Wild Audio and Spelling Variations
Sep 21, 2026Kaushal Santosh Bhogale, Srija Anand, Sadakopa Ramakrishnan Thothathiri +3ASR Evaluation Demographic Bias in ASR
Indian Institute of Technology, Madras, India · Sarvam AI, India
A Probe Shift Is Not a Fairness Fix: The Limits of Representation Steering in Speech Models
Sep 16, 2026Nicolas Bourrel, Abderrahmane Issam, Gerasimos SpanakisASR Evaluation Demographic Bias in ASR
Department of Advanced Computing Sciences Maastricht University
SamaVaani: Auditing and Debiasing Multilingual Clinical ASR for Indian Languages
Jun 25, 2026Subham Kumar, Prakrithi Shivaprakash, Abhishek Manoharan +7ASR Evaluation Demographic Bias in ASR
IIT, Kharagpur · NIMHANS, Bangalore · LGBRIMH, Tezpur
ROMPAR: Morphological Completion and Demographic Unlearning for Romanian-Accented Speech Recognition
Jun 14, 2026Andrei-Marius Avram, Aureliu-Valentin Antonie, Ştefan-Bogdan Badea +3Adversarial Training Demographic Bias in ASR
National University of Science and Technology POLITEHNICA Bucharest, Romania
Evaluating Bias in Phoneme-Based Automatic Speech Recognition Systems: An Analysis of IPA Transcription Models
Jun 10, 2026Maneesha Rani Saha, Catherine Bao, Neal PatwariASR Evaluation Demographic Bias in ASR
University of Utah
Speaker Group Encoding in Self-supervised Speech Recognition Models
Jun 9, 2026Felix Herron, Solange Rossato Alexandre Allauzen, Benoit Favre +1Audio Representation Learning Demographic Bias in ASR
MILES Team, LAMSADE, Université Paris Dauphine-PSL, France (1) · GETALP Team, LIG, Université Grenoble Alpes, France (2) · NLP team, LIS, Aix-Marseille University, France (3)
Your Multimodal Speech Model Says I Have a Face for Radio
May 28, 2026Maya K. Nachesa, Vlad Niculae, Vagrant GautamASR Evaluation Demographic Bias in ASR
Language Technology Lab University of Amsterdam · Heidelberg Institute for Theoretical Studies
Responsible Benchmarking of Fairness for Automatic Speech Recognition
May 11, 2026Felix Herron, Ange Richard, François Portet +2Algorithmic Fairness Demographic Bias in ASR
MILES Team, LAMSADE, Université Paris Dauphine-PSL (1) · GETALP Team, LIG, Université Grenoble Alpes (2) · PACTE, Université Grenoble-Alpes (3)
Identifying and typifying demographic unfairness in phoneme-level embeddings of self-supervised speech recognition models
Apr 24, 2026Felix Herron, Solange Rossato, Alexandre Allauzen +1Demographic Bias in ASR Automatic Speech Recognition
MILES Team, LAMSADE, Université Paris Dauphine-PSL, France · GETALP Team, LIG, Université Grenoble Alpes, France
Do LLM Decoders Listen Fairly? Benchmarking How Language Model Priors Shape Bias in Speech Recognition
Apr 23, 2026Srishti Ginjala, Eric Fosler-Lussier, Christopher W. Myers +1ASR Evaluation Demographic Bias in ASR
The Ohio State University, Columbus, OH, USA · Air Force Research Laboratory, USA
"This Wasn't Made for Me": Recentering User Experience and Emotional Impact in the Evaluation of ASR Bias
Apr 22, 2026Siyu Liang, Alicia Beckford WassinkAlgorithmic Fairness Demographic Bias in ASR
University of Washington, Seattle, Washington, USA
Where Do Self-Supervised Speech Models Become Unfair?
Apr 20, 2026Felix Herron, Maja Hjuler, Solange Rossato +2Speech Processing Demographic Bias in ASR
MILES Team, LAMSADE, Universit´e Paris Dauphine-PSL · GETALP Team, LIG, Universit´e Grenoble Alpes · Queensland University of Technology

Oct 6, 2026, cs.CLJ/K move · Enter open · S save

InterCorrect: Intersection-Aware Correction of Demographic Model Merging for Fair ASR

Ashley E. Bravo-Bravo, Yuchen Zhang, Haralambos Mouratidis, Ravi Shekhar+1

PUCP · IADS, University of Essex · CSEE, University of Essex+1

Demographic Bias in ASR Model Merging Speech Language Models

Open paper AskPDF arXiv

Automatic Speech Recognition (ASR) systems often show uneven performance across demographic groups, and errors can be especially difficult to address for speakers belonging to multiple demographic groups. This work studies demographic-aware model merging for fair Speech-LLM-based ASR. Starting from a SLAM-ASR-based model, we fine-tune only the connector on demographic-specific subsets and merge the resulting subgroup-adapted connectors into a global model. We then identify critical cross-axis demographic pairs using subgroup WER and task-vector conflict, and apply intersection-specific correction vectors to the global merged model. Experiments on Fair-Speech show that global demographic merging improves overall WER over the base model, while intersection correction provides additional gains for several merging strategies. In particular, TIES with WER-based correction achieves the best overall WER, reducing it from 7.38% to 5.13%. Subgroup and disparity analyses further show that the proposed approach improves performance across demographic axes, while highlighting that lower average WER does not always imply reduced subgroup disparity.

Momentum

Related topics

Latest papers 16