English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

Machine learning fairness concerns about the biases towards certain protectedor sensitive group of people when addressing the target tasks. This paperstudies the debiasing problem in the context of image classification tasks. Ourdata analysis on facial attribute recognition demonstrates (1) the attributionof model bias from imbalanced training data distribution and (2) the potentialof adversarial examples in balancing data distribution. We are thus motivatedto employ adversarial example to augment the training data for visualdebiasing. Specifically, to ensure the adversarial generalization as well ascross-task transferability, we propose to couple the operations of target taskclassifier training, bias task classifier training, and adversarial examplegeneration. The generated adversarial examples supplement the target tasktraining dataset via balancing the distribution over bias variables in anonline fashion. Results on simulated and real-world debiasing experimentsdemonstrate the effectiveness of the proposed solution in simultaneouslyimproving model accuracy and fairness. Preliminary experiment on few-shotlearning further shows the potential of adversarial attack-based pseudo samplegeneration as alternative solution to make up for the training data lackage.

Towards Accuracy-Fairness Paradox: Adversarial Example-based Data Augmentation for Visual Debiasing