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

Face anti-spoofing (FAS) plays a critical role in securing face recognitionsystems from different presentation attacks. Previous works leverage auxiliarypixel-level supervision and domain generalization approaches to address unseenspoof types. However, the local characteristics of image captures, i.e.,capturing devices and presenting materials, are ignored in existing works andwe argue that such information is required for networks to discriminate betweenlive and spoof images. In this work, we propose PatchNet which reformulatesface anti-spoofing as a fine-grained patch-type recognition problem. To bespecific, our framework recognizes the combination of capturing devices andpresenting materials based on the patches cropped from non-distorted faceimages. This reformulation can largely improve the data variation and enforcethe network to learn discriminative feature from local capture patterns. Inaddition, to further improve the generalization ability of the spoof feature,we propose the novel Asymmetric Margin-based Classification Loss andSelf-supervised Similarity Loss to regularize the patch embedding space. Ourexperimental results verify our assumption and show that the model is capableof recognizing unseen spoof types robustly by only looking at local regions.Moreover, the fine-grained and patch-level reformulation of FAS outperforms theexisting approaches on intra-dataset, cross-dataset, and domain generalizationbenchmarks. Furthermore, our PatchNet framework can enable practicalapplications like Few-Shot Reference-based FAS and facilitate futureexploration of spoof-related intrinsic cues.

PatchNet: A Simple Face Anti-Spoofing Framework via Fine-Grained Patch Recognition