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

Vision-language pre-training models (VLPs) have exhibited revolutionaryimprovements in various vision-language tasks. In VLP, some adversarial attacksfool a model into false or absurd classifications. Previous studies addressedthese attacks by fine-tuning the model or changing its architecture. However,these methods risk losing the original model's performance and are difficult toapply to downstream tasks. In particular, their applicability to other taskshas not been considered. In this study, we addressed the reduction of theimpact of typographic attacks on CLIP without changing the model parameters. Toachieve this, we expand the idea of "prefix learning" and introduce our simpleyet effective method: Defense-Prefix (DP), which inserts the DP token before aclass name to make words "robust" against typographic attacks. Our method canbe easily applied to downstream tasks, such as object detection, because theproposed method is independent of the model parameters. Our methodsignificantly improves the accuracy of classification tasks for typographicattack datasets, while maintaining the zero-shot capabilities of the model. Inaddition, we leverage our proposed method for object detection, demonstratingits high applicability and effectiveness. The codes and datasets are availableat https://github.com/azuma164/Defense-Prefix.

Defense-Prefix for Preventing Typographic Attacks on CLIP