Towards Fairness in Visual Recognition: Effective Strategies for Bias Mitigation

Abstract

Computer vision models learn to perform a task by capturing relevant statistics from training data. It has been shown that models learn spurious age, gender, and race correlations when trained for seemingly unrelated tasks like activity recognition or image captioning. Various mitigation techniques have been presented to prevent models from utilizing or learning such biases. However, there has been little systematic comparison between these techniques. We design a simple but surprisingly effective visual recognition benchmark for studying bias mitigation. Using this benchmark, we provide a thorough analysis of a wide range of techniques. We highlight the shortcomings of popular adversarial training approaches for bias mitigation, propose a simple but similarly effective alternative to the inference-time Reducing Bias Amplification method of Zhao et al., and design a domain-independent training technique that outperforms all other methods. Finally, we validate our findings on the attribute classification task in the CelebA dataset, where attribute presence is known to be correlated with the gender of people in the image, and demonstrate that the proposed technique is effective at mitigating real-world gender bias.

arXiv commentTo appear in CVPR 2020

References in corpus (13)

• A systematic study of the class imbalance problem in convolutional neural networks
• Semantics derived automatically from language corpora contain human-like biases
• Equality of Opportunity in Supervised Learning
• Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings
• What makes ImageNet good for transfer learning?
• On Fairness and Calibration
• On Formalizing Fairness in Prediction with Machine Learning
• Flexibly Fair Representation Learning by Disentanglement
• Women also Snowboard: Overcoming Bias in Captioning Models
• InclusiveFaceNet: Improving Face Attribute Detection with Race and Gender Diversity
• Bias Correction of Learned Generative Models using Likelihood-Free Importance Weighting
• Fair Generative Modeling via Weak Supervision
• Feature-Wise Bias Amplification

Cited by in corpus (3)

• A Survey of Deep Learning for Scientific Discovery
• Uncovering the Bias in Facial Expressions
• Dataset Bias Mitigation Through Analysis of CNN Training Scores