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/

Global: Zendy Plus annual

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

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Contextual information is a valuable cue for Deep Neural Networks (DNNs) tolearn better representations and improve accuracy. However, co-occurrence biasin the training dataset may hamper a DNN model's generalizability to unseenscenarios in the real world. For example, in COCO, many object categories havea much higher co-occurrence with men compared to women, which can bias a DNN'sprediction in favor of men. Recent works have focused on task-specific trainingstrategies to handle bias in such scenarios, but fixing the available data isoften ignored. In this paper, we propose a novel and more generic solution toaddress the contextual bias in the datasets by selecting a subset of thesamples, which is fair in terms of the co-occurrence with various classes for aprotected attribute. We introduce a data repair algorithm using the coefficientof variation, which can curate fair and contextually balanced data for aprotected class(es). This helps in training a fair model irrespective of thetask, architecture or training methodology. Our proposed solution is simple,effective, and can even be used in an active learning setting where the datalabels are not present or being generated incrementally. We demonstrate theeffectiveness of our algorithm for the task of object detection and multi-labelimage classification across different datasets. Through a series ofexperiments, we validate that curating contextually fair data helps make modelpredictions fair by balancing the true positive rate for the protected classacross groups without compromising on the model's overall performance.

Does Data Repair Lead to Fair Models? Curating Contextually Fair Data To Reduce Model Bias