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

For stability and reliability of real-world applications, the robustness ofDNNs in unimodal tasks has been evaluated. However, few studies considerabnormal situations that a visual question answering (VQA) model mightencounter at test time after deployment in the real-world. In this study, weevaluate the robustness of state-of-the-art VQA models to five differentanomalies, including worst-case scenarios, the most frequent scenarios, and thecurrent limitation of VQA models. Different from the results in unimodal tasks,the maximum confidence of answers in VQA models cannot detect anomalous inputs,and post-training of the outputs, such as outlier exposure, is ineffective forVQA models. Thus, we propose an attention-based method, which uses confidenceof reasoning between input images and questions and shows much more promisingresults than the previous methods in unimodal tasks. In addition, we show thata maximum entropy regularization of attention networks can significantlyimprove the attention-based anomaly detection of the VQA models. Thanks to thesimplicity, attention-based anomaly detection and the regularization aremodel-agnostic methods, which can be used for various cross-modal attentions inthe state-of-the-art VQA models. The results imply that cross-modal attentionin VQA is important to improve not only VQA accuracy, but also the robustnessto various anomalies.

Regularizing Attention Networks for Anomaly Detection in Visual Question Answering