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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Recent studies have shown that higher accuracy on ImageNet usually leads tobetter robustness against different corruptions. Therefore, in this paper,instead of following the traditional research paradigm that investigates newout-of-distribution corruptions or perturbations deep models may encounter, weconduct model debugging in in-distribution data to explore which objectattributes a model may be sensitive to. To achieve this goal, we create atoolkit for object editing with controls of backgrounds, sizes, positions, anddirections, and create a rigorous benchmark named ImageNet-E(diting) forevaluating the image classifier robustness in terms of object attributes. Withour ImageNet-E, we evaluate the performance of current deep learning models,including both convolutional neural networks and vision transformers. We findthat most models are quite sensitive to attribute changes. A small change inthe background can lead to an average of 9.23\% drop on top-1 accuracy. We alsoevaluate some robust models including both adversarially trained models andother robust trained models and find that some models show worse robustnessagainst attribute changes than vanilla models. Based on these findings, wediscover ways to enhance attribute robustness with preprocessing, architecturedesigns, and training strategies. We hope this work can provide some insightsto the community and open up a new avenue for research in robust computervision. The code and dataset are available athttps://github.com/alibaba/easyrobust.

ImageNet-E: Benchmarking Neural Network Robustness via Attribute Editing