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

Recent studies have shown that neural network (NN) based image classifiersare highly vulnerable to adversarial examples, which poses a threat tosecurity-sensitive image recognition task. Prior work has shown that JPEGcompression can combat the drop in classification accuracy on adversarialexamples to some extent. But, as the compression ratio increases, traditionalJPEG compression is insufficient to defend those attacks but can cause anabrupt accuracy decline to the benign images. In this paper, with the aim offully filtering the adversarial perturbations, we firstly make modifications totraditional JPEG compression algorithm which becomes more favorable for NN.Specifically, based on an analysis of the frequency coefficient, we design aNN-favored quantization table for compression. Considering compression as adata augmentation strategy, we then combine our model-agnostic preprocess withnoisy training. We fine-tune the pre-trained model by training with imagesencoded at different compression levels, thus generating multiple classifiers.Finally, since lower (higher) compression ratio can remove both perturbationsand original features slightly (aggressively), we use these trained multiplemodels for model ensemble. The majority vote of the ensemble of models isadopted as final predictions. Experiments results show our method can improvedefense efficiency while maintaining original accuracy.

Countering Adversarial Examples: Combining Input Transformation and Noisy Training