# Learning Bias-Invariant Representation by Cross-Sample Mutual Information Minimization Source: [Learning Bias-Invariant Representation by Cross-Sample Mutual Information Minimization]() Collection index: [llms.txt]() Data updated: 2025-10-27T06:09:35.928Z Authors: Wei Zhu, Haitian Zheng, Haofu Liao, Weijian Li, Jiebo Luo Published: 2021-08-11T21:17:02.000Z ## Abstract Deep learning algorithms mine knowledge from the training data and thus would likely inherit the dataset's bias information. As a result, the obtained model would generalize poorly and even mislead the decision process in real-life applications. We propose to remove the bias information misused by the target task with a cross-sample adversarial debiasing (CSAD) method. CSAD explicitly extracts target and bias features disentangled from the latent representation generated by a feature extractor and then learns to discover and remove the correlation between the target and bias features. The correlation measurement plays a critical role in adversarial debiasing and is conducted by a cross-sample neural mutual information estimator. Moreover, we propose joint content and local structural representation learning to boost mutual information estimation for better performance. We conduct thorough experiments on publicly available datasets to validate the advantages of the proposed method over state-of-the-art approaches. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Image Classification on CelebA HairColor](): Image Classification; CelebA HairColor; Guiding Score: 91.19. - [Image Classification on CelebA HeavyMakeup](): Image Classification; CelebA HeavyMakeup; Guiding Score: 0.8232. ## Original sources and implementation [Original paper]() [PDF]()