Matches in SemOpenAlex for { <https://semopenalex.org/work/W4304080557> ?p ?o ?g. }
- W4304080557 abstract "The deep learning-based face forgery detection is a novel yet challenging task. Despite impressive results have been achieved, there are still some limitations in the existing methods. For example, the previous methods are hard to maintain consistent predictions for consecutive frames, even if all of those frames are actually forged. We propose a symmetric transformer for channel and spatial feature extraction, which is because the channel and spatial features of a robust forgery detector should be consistent in the temporal domain. The symmetric transformer adopt the newly-designed attention-based strategies for channel variance and spatial gradients as the vital features, which greatly improves the robustness of deepfake video detection. Moreover, this symmetric structure acts on temporal and spatial features respectively, which ensures the robustness of detection from two different aspects. Our symmetric transformer is an end-to-end optimized network. Experiments are conducted on various settings, the proposed methods achieve significantly improvement on prediction robustness and perform better than state-of-the-art methods on different datasets." @default.
- W4304080557 created "2022-10-10" @default.
- W4304080557 creator A5024413169 @default.
- W4304080557 creator A5024920256 @default.
- W4304080557 creator A5058022124 @default.
- W4304080557 creator A5064805926 @default.
- W4304080557 creator A5071690624 @default.
- W4304080557 creator A5071772000 @default.
- W4304080557 date "2022-10-10" @default.
- W4304080557 modified "2023-10-14" @default.
- W4304080557 title "Face Forgery Detection via Symmetric Transformer" @default.
- W4304080557 cites W1522734439 @default.
- W4304080557 cites W1677182931 @default.
- W4304080557 cites W1968164921 @default.
- W4304080557 cites W2009130368 @default.
- W4304080557 cites W2071794886 @default.
- W4304080557 cites W2095613088 @default.
- W4304080557 cites W2412509443 @default.
- W4304080557 cites W2486034530 @default.
- W4304080557 cites W2507009361 @default.
- W4304080557 cites W2531409750 @default.
- W4304080557 cites W2603123944 @default.
- W4304080557 cites W2738947046 @default.
- W4304080557 cites W2752782242 @default.
- W4304080557 cites W2786289897 @default.
- W4304080557 cites W2891145043 @default.
- W4304080557 cites W2913399670 @default.
- W4304080557 cites W2942074357 @default.
- W4304080557 cites W2955058313 @default.
- W4304080557 cites W2962934715 @default.
- W4304080557 cites W2962958939 @default.
- W4304080557 cites W2963524571 @default.
- W4304080557 cites W2963684180 @default.
- W4304080557 cites W2982058372 @default.
- W4304080557 cites W2990503944 @default.
- W4304080557 cites W2997150500 @default.
- W4304080557 cites W3034175346 @default.
- W4304080557 cites W3034196597 @default.
- W4304080557 cites W3034577585 @default.
- W4304080557 cites W3034713808 @default.
- W4304080557 cites W3034785167 @default.
- W4304080557 cites W3034795015 @default.
- W4304080557 cites W3034900344 @default.
- W4304080557 cites W3035063907 @default.
- W4304080557 cites W3036576316 @default.
- W4304080557 cites W3096609285 @default.
- W4304080557 cites W3205994442 @default.
- W4304080557 cites W4312593844 @default.
- W4304080557 doi "https://doi.org/10.1145/3503161.3547806" @default.
- W4304080557 hasPublicationYear "2022" @default.
- W4304080557 type Work @default.
- W4304080557 citedByCount "3" @default.
- W4304080557 countsByYear W43040805572023 @default.
- W4304080557 crossrefType "proceedings-article" @default.
- W4304080557 hasAuthorship W4304080557A5024413169 @default.
- W4304080557 hasAuthorship W4304080557A5024920256 @default.
- W4304080557 hasAuthorship W4304080557A5058022124 @default.
- W4304080557 hasAuthorship W4304080557A5064805926 @default.
- W4304080557 hasAuthorship W4304080557A5071690624 @default.
- W4304080557 hasAuthorship W4304080557A5071772000 @default.
- W4304080557 hasConcept C104317684 @default.
- W4304080557 hasConcept C119599485 @default.
- W4304080557 hasConcept C127413603 @default.
- W4304080557 hasConcept C153180895 @default.
- W4304080557 hasConcept C154945302 @default.
- W4304080557 hasConcept C165801399 @default.
- W4304080557 hasConcept C185592680 @default.
- W4304080557 hasConcept C31972630 @default.
- W4304080557 hasConcept C41008148 @default.
- W4304080557 hasConcept C52622490 @default.
- W4304080557 hasConcept C55493867 @default.
- W4304080557 hasConcept C63479239 @default.
- W4304080557 hasConcept C66322947 @default.
- W4304080557 hasConcept C76155785 @default.
- W4304080557 hasConcept C94915269 @default.
- W4304080557 hasConceptScore W4304080557C104317684 @default.
- W4304080557 hasConceptScore W4304080557C119599485 @default.
- W4304080557 hasConceptScore W4304080557C127413603 @default.
- W4304080557 hasConceptScore W4304080557C153180895 @default.
- W4304080557 hasConceptScore W4304080557C154945302 @default.
- W4304080557 hasConceptScore W4304080557C165801399 @default.
- W4304080557 hasConceptScore W4304080557C185592680 @default.
- W4304080557 hasConceptScore W4304080557C31972630 @default.
- W4304080557 hasConceptScore W4304080557C41008148 @default.
- W4304080557 hasConceptScore W4304080557C52622490 @default.
- W4304080557 hasConceptScore W4304080557C55493867 @default.
- W4304080557 hasConceptScore W4304080557C63479239 @default.
- W4304080557 hasConceptScore W4304080557C66322947 @default.
- W4304080557 hasConceptScore W4304080557C76155785 @default.
- W4304080557 hasConceptScore W4304080557C94915269 @default.
- W4304080557 hasLocation W43040805571 @default.
- W4304080557 hasOpenAccess W4304080557 @default.
- W4304080557 hasPrimaryLocation W43040805571 @default.
- W4304080557 hasRelatedWork W1602592726 @default.
- W4304080557 hasRelatedWork W1999222583 @default.
- W4304080557 hasRelatedWork W2035976912 @default.
- W4304080557 hasRelatedWork W2042311553 @default.
- W4304080557 hasRelatedWork W2109974539 @default.
- W4304080557 hasRelatedWork W2125927971 @default.
- W4304080557 hasRelatedWork W2541791370 @default.