CO-VADA: A Confidence-Oriented Voice Augmentation Debiasing Approach for Fair Speech Emotion Recognition
Not on the map. Why?
Abstract (arXiv)
Bias in speech emotion recognition (SER) systems often stems from spurious correlations between speaker characteristics and emotional labels, leading to unfair predictions across demographic groups. Many existing debiasing methods require model-specific changes or demographic annotations, limiting their practical use. We present CO-VADA, a Confidence-Oriented Voice Augmentation Debiasing Approach that mitigates bias without modifying model architecture or relying on demographic information. CO-VADA identifies training samples that reflect bias patterns present in the training data and then applies voice conversion to alter irrelevant attributes and generate samples. These augmented samples introduce speaker variations that differ from dominant patterns in the data, guiding the model to focus more on emotion-relevant features. Our framework is compatible with various SER models and voice conversion tools, making it a scalable and practical solution for improving fairness in SER systems.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 39 of its 65 references that are in the map. First posted on arXiv 6 Jun 2025.
Builds on
39 of 65 references.
References · 39
- 2016Equality of Opportunity in Supervised Learningcited by 2,007
- 2021WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processingcited by 1,657
- 2021SUPERB: Speech processing Universal PERformance Benchmarkcited by 666
- 2021Just Train Twice: Improving Group Robustness without Training Group Informationcited by 523
- 2019End-to-End Bias Mitigation by Modelling Biases in Corporacited by 128
- 2021Learning Debiased Representation via Disentangled Feature Augmentationcited by 116
- 2022FreeVC: Towards High-Quality Text-Free One-Shot Voice Conversioncited by 80
- 2021BiaSwap: Removing dataset bias with bias-tailored swapping augmentationcited by 70
- 2023Voice Conversion With Just Nearest Neighborscited by 68
- 2018Domain Adversarial Training for Accented Speech Recognitioncited by 40
- 2021Balancing out Bias: Achieving Fairness Through Balanced Trainingcited by 34
- 2023Diff-HierVC: Diffusion-based Hierarchical Voice Conversion with Robust Pitch Generation and Masked Prior for Zero-shot Speaker Adaptationcited by 30
27 more
- 2022Toward Fairness in Speech Recognition: Discovery and mitigation of performance disparitiescited by 28
- 2023Kid-Whisper: Towards Bridging the Performance Gap in Automatic Speech Recognition for Children VS. Adultscited by 27
- 2022BLIND: Bias Removal With No Demographicscited by 22
- 2024Listen and Speak Fairly: A Study on Semantic Gender Bias in Speech Integrated Large Language Modelscited by 21
- 2021Don't speak too fast: The impact of data bias on self-supervised speech modelscited by 20
- 2024Emo-bias: A Large Scale Evaluation of Social Bias on Speech Emotion Recognitioncited by 20
- 2021Adversarial Data Augmentation for Disordered Speech Recognitioncited by 18
- 2024On the social bias of speech self-supervised modelscited by 16
- 2024Spoken Stereoset: On Evaluating Social Bias Toward Speaker in Speech Large Language Modelscited by 16
- 2022Speech Emotion Recognition using Supervised Deep Recurrent System for Mental Health Monitoringcited by 15
- 2022Improving fairness in speaker verification via Group-adapted Fusion Networkcited by 11
- 2023Improving Fairness and Robustness in End-to-End Speech Recognition through unsupervised clusteringcited by 11
- 2023Exploring data augmentation in bias mitigation against non-native-accented speechcited by 11
- 2024Improving Speech Emotion Recognition in Under-Resourced Languages via Speech-to-Speech Translation with Bootstrapping Data Selectioncited by 10
- 2023Pre-trained Speech Processing Models Contain Human-Like Biases that Propagate to Speech Emotion Recognitioncited by 9
- 2024A Contrastive Learning Approach to Mitigate Bias in Speech Modelscited by 9
- 2023Data Augmentation via Subgroup Mixup for Improving Fairnesscited by 8
- 2023End-to-End Continuous Speech Emotion Recognition in Real-life Customer Service Call Center Conversationscited by 6
- 2025Mitigating Subgroup Disparities in Multi-Label Speech Emotion Recognition: A Pseudo-Labeling and Unsupervised Learning Approachcited by 5
- 2023Testing Correctness, Fairness, and Robustness of Speech Emotion Recognition Modelscited by 4
- 2024Addressing Both Statistical and Causal Gender Fairness in NLP Modelscited by 4
- 2023Debiased Automatic Speech Recognition for Dysarthric Speech via Sample Reweighting with Sample Affinity Testcited by 3
- 2023Signal Is Harder To Learn Than Bias: Debiasing with Focal Losscited by 3
- 2024Unlabeled Debiasing in Downstream Tasks via Class-wise Low Variance Regularizationcited by 3
- 2025Is It Still Fair? Investigating Gender Fairness in Cross-Corpus Speech Emotion Recognitioncited by 3
- 2025Domain Adversarial Training for Mitigating Gender Bias in Speech-based Mental Health Detectioncited by 3
- 2022Adversarial Reweighting for Speaker Verification Fairnesscited by 2
Built on by
Citing papers per year: 2026: 1.
1 paper.