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

In the domain of Visual Question Answering (VQA), studies have shownimprovement in users' mental model of the VQA system when they are exposed toexamples of how these systems answer certain Image-Question (IQ) pairs. In thiswork, we show that showing controlled counterfactual image-question examplesare more effective at improving the mental model of users as compared to simplyshowing random examples. We compare a generative approach and a retrieval-basedapproach to show counterfactual examples. We use recent advances in generativeadversarial networks (GANs) to generate counterfactual images by deleting andinpainting certain regions of interest in the image. We then expose users tochanges in the VQA system's answer on those altered images. To select theregion of interest for inpainting, we experiment with using bothhuman-annotated attention maps and a fully automatic method that uses the VQAsystem's attention values. Finally, we test the user's mental model by askingthem to predict the model's performance on a test counterfactual image. We notean overall improvement in users' accuracy to predict answer change when showncounterfactual explanations. While realistic retrieved counterfactualsobviously are the most effective at improving the mental model, we show that agenerative approach can also be equally effective.

Improving Users' Mental Model with Attention-directed Counterfactual Edits