Matches in SemOpenAlex for { <https://semopenalex.org/work/W4312906857> ?p ?o ?g. }
- W4312906857 abstract "Deep spatiotemporal models are used in a variety of computer vision tasks, such as action recognition and video object segmentation. Currently, there is a limited understanding of what information is captured by these models in their intermediate representations. For example, while it has been observed that action recognition algorithms are heavily influenced by visual appearance in single static frames, there is no quantitative methodology for evaluating such static bias in the latent representation compared to bias toward dynamic information (e.g. motion). We tackle this challenge by proposing a novel approach for quantifying the static and dynamic biases of any spatiotemporal model. To show the efficacy of our approach, we analyse two widely studied tasks, action recognition and video object segmentation. Our key findings are threefold: (i) Most examined spatiotemporal models are biased toward static information; although, certain two-stream architectures with cross-connections show a better balance between the static and dynamic information captured. (ii) Some datasets that are commonly assumed to be biased toward dynamics are actually biased toward static information. (iii) Individual units (channels) in an architecture can be biased toward static, dynamic or a combination of the two. <sup xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>1</sup> <sup xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>1</sup> Project page and code" @default.
- W4312906857 created "2023-01-05" @default.
- W4312906857 creator A5022335103 @default.
- W4312906857 creator A5033145377 @default.
- W4312906857 creator A5052269532 @default.
- W4312906857 creator A5072076034 @default.
- W4312906857 creator A5087848827 @default.
- W4312906857 creator A5089232277 @default.
- W4312906857 date "2022-06-01" @default.
- W4312906857 modified "2023-09-25" @default.
- W4312906857 title "A Deeper Dive Into What Deep Spatiotemporal Networks Encode: Quantifying Static vs. Dynamic Information" @default.
- W4312906857 cites W1522734439 @default.
- W4312906857 cites W1983364832 @default.
- W4312906857 cites W2073982870 @default.
- W4312906857 cites W2092939357 @default.
- W4312906857 cites W2111045798 @default.
- W4312906857 cites W2141859737 @default.
- W4312906857 cites W2194775991 @default.
- W4312906857 cites W2470139095 @default.
- W4312906857 cites W259727857 @default.
- W4312906857 cites W2603777577 @default.
- W4312906857 cites W2610147486 @default.
- W4312906857 cites W2625366777 @default.
- W4312906857 cites W2896347987 @default.
- W4312906857 cites W2963091558 @default.
- W4312906857 cites W2963524571 @default.
- W4312906857 cites W2963616706 @default.
- W4312906857 cites W2982349016 @default.
- W4312906857 cites W2990503944 @default.
- W4312906857 cites W2997487053 @default.
- W4312906857 cites W3034572008 @default.
- W4312906857 cites W3048537404 @default.
- W4312906857 cites W3108460979 @default.
- W4312906857 cites W3118473641 @default.
- W4312906857 cites W3170511209 @default.
- W4312906857 cites W3190415286 @default.
- W4312906857 cites W4214614183 @default.
- W4312906857 cites W4288083805 @default.
- W4312906857 doi "https://doi.org/10.1109/cvpr52688.2022.01361" @default.
- W4312906857 hasPublicationYear "2022" @default.
- W4312906857 type Work @default.
- W4312906857 citedByCount "6" @default.
- W4312906857 countsByYear W43129068572022 @default.
- W4312906857 countsByYear W43129068572023 @default.
- W4312906857 crossrefType "proceedings-article" @default.
- W4312906857 hasAuthorship W4312906857A5022335103 @default.
- W4312906857 hasAuthorship W4312906857A5033145377 @default.
- W4312906857 hasAuthorship W4312906857A5052269532 @default.
- W4312906857 hasAuthorship W4312906857A5072076034 @default.
- W4312906857 hasAuthorship W4312906857A5087848827 @default.
- W4312906857 hasAuthorship W4312906857A5089232277 @default.
- W4312906857 hasBestOaLocation W43129068572 @default.
- W4312906857 hasConcept C104317684 @default.
- W4312906857 hasConcept C11413529 @default.
- W4312906857 hasConcept C119857082 @default.
- W4312906857 hasConcept C121332964 @default.
- W4312906857 hasConcept C153180895 @default.
- W4312906857 hasConcept C154945302 @default.
- W4312906857 hasConcept C177264268 @default.
- W4312906857 hasConcept C17744445 @default.
- W4312906857 hasConcept C185592680 @default.
- W4312906857 hasConcept C199360897 @default.
- W4312906857 hasConcept C199539241 @default.
- W4312906857 hasConcept C26517878 @default.
- W4312906857 hasConcept C2776359362 @default.
- W4312906857 hasConcept C2776760102 @default.
- W4312906857 hasConcept C2780791683 @default.
- W4312906857 hasConcept C2781238097 @default.
- W4312906857 hasConcept C31972630 @default.
- W4312906857 hasConcept C37404715 @default.
- W4312906857 hasConcept C38652104 @default.
- W4312906857 hasConcept C41008148 @default.
- W4312906857 hasConcept C55493867 @default.
- W4312906857 hasConcept C62520636 @default.
- W4312906857 hasConcept C66746571 @default.
- W4312906857 hasConcept C89600930 @default.
- W4312906857 hasConcept C94625758 @default.
- W4312906857 hasConceptScore W4312906857C104317684 @default.
- W4312906857 hasConceptScore W4312906857C11413529 @default.
- W4312906857 hasConceptScore W4312906857C119857082 @default.
- W4312906857 hasConceptScore W4312906857C121332964 @default.
- W4312906857 hasConceptScore W4312906857C153180895 @default.
- W4312906857 hasConceptScore W4312906857C154945302 @default.
- W4312906857 hasConceptScore W4312906857C177264268 @default.
- W4312906857 hasConceptScore W4312906857C17744445 @default.
- W4312906857 hasConceptScore W4312906857C185592680 @default.
- W4312906857 hasConceptScore W4312906857C199360897 @default.
- W4312906857 hasConceptScore W4312906857C199539241 @default.
- W4312906857 hasConceptScore W4312906857C26517878 @default.
- W4312906857 hasConceptScore W4312906857C2776359362 @default.
- W4312906857 hasConceptScore W4312906857C2776760102 @default.
- W4312906857 hasConceptScore W4312906857C2780791683 @default.
- W4312906857 hasConceptScore W4312906857C2781238097 @default.
- W4312906857 hasConceptScore W4312906857C31972630 @default.
- W4312906857 hasConceptScore W4312906857C37404715 @default.
- W4312906857 hasConceptScore W4312906857C38652104 @default.
- W4312906857 hasConceptScore W4312906857C41008148 @default.
- W4312906857 hasConceptScore W4312906857C55493867 @default.
- W4312906857 hasConceptScore W4312906857C62520636 @default.
- W4312906857 hasConceptScore W4312906857C66746571 @default.