FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386076265> ?p ?o ?g. }

• W4386076265 abstract "Contrastive learning-based video-language representation learning approaches, e.g., CLIP, have achieved outstanding performance, which pursue semantic interaction upon pre-defined video-text pairs. To clarify this coarse-grained global interaction and move a step further, we have to encounter challenging shell-breaking interactions for fine-grained cross-modal learning. In this paper, we creatively model video-text as game players with multivariate cooperative game theory to wisely handle the uncertainty during fine-grained semantic interaction with diverse granularity, flexible combination, and vague intensity. Concretely, we propose Hierarchical Banzhaf Interaction (HBI) to value possible correspondence between video frames and text words for sensitive and explainable cross-modal contrast. To efficiently realize the cooperative game of multiple video frames and multiple text words, the proposed method clusters the original video frames (text words) and computes the Banzhaf Interaction between the merged tokens. By stacking token merge modules, we achieve cooperative games at different semantic levels. Extensive experiments on commonly used text-video retrieval and video-question answering bench-marks with superior performances justify the efficacy of our HBI. More encouragingly, it can also serve as a visualization tool to promote the understanding of cross-modal interaction, which have a far-reaching impact on the community. Project page is available at https://jpthu17.github.io/HBI/." @default.
• W4386076265 created "2023-08-23" @default.
• W4386076265 creator A5001679038 @default.
• W4386076265 creator A5010479652 @default.
• W4386076265 creator A5019381882 @default.
• W4386076265 creator A5022256556 @default.
• W4386076265 creator A5024401174 @default.
• W4386076265 creator A5057135582 @default.
• W4386076265 creator A5064414647 @default.
• W4386076265 creator A5065035829 @default.
• W4386076265 date "2023-06-01" @default.
• W4386076265 modified "2023-10-06" @default.
• W4386076265 title "Video-Text as Game Players: Hierarchical Banzhaf Interaction for Cross-Modal Representation Learning" @default.
• W4386076265 cites W2009662043 @default.
• W4386076265 cites W2038127015 @default.
• W4386076265 cites W2091836314 @default.
• W4386076265 cites W2103691541 @default.
• W4386076265 cites W2268194897 @default.
• W4386076265 cites W2425121537 @default.
• W4386076265 cites W2510508396 @default.
• W4386076265 cites W2765716052 @default.
• W4386076265 cites W2963017553 @default.
• W4386076265 cites W2963916161 @default.
• W4386076265 cites W3035356601 @default.
• W4386076265 cites W3035524453 @default.
• W4386076265 cites W3121356232 @default.
• W4386076265 cites W3128401049 @default.
• W4386076265 cites W3145807616 @default.
• W4386076265 cites W3153005511 @default.
• W4386076265 cites W3167092180 @default.
• W4386076265 cites W3168640669 @default.
• W4386076265 cites W3169472988 @default.
• W4386076265 cites W3174873881 @default.
• W4386076265 cites W3175888430 @default.
• W4386076265 cites W3197457832 @default.
• W4386076265 cites W3204588463 @default.
• W4386076265 cites W3204670646 @default.
• W4386076265 cites W4211220176 @default.
• W4386076265 cites W4285606530 @default.
• W4386076265 cites W4290874851 @default.
• W4386076265 cites W4312372711 @default.
• W4386076265 cites W4312530435 @default.
• W4386076265 cites W4312761939 @default.
• W4386076265 cites W4312991156 @default.
• W4386076265 cites W4312999114 @default.
• W4386076265 cites W4313196315 @default.
• W4386076265 doi "https://doi.org/10.1109/cvpr52729.2023.00244" @default.
• W4386076265 hasPublicationYear "2023" @default.
• W4386076265 type Work @default.
• W4386076265 citedByCount "1" @default.
• W4386076265 countsByYear W43860762652023 @default.
• W4386076265 crossrefType "proceedings-article" @default.
• W4386076265 hasAuthorship W4386076265A5001679038 @default.
• W4386076265 hasAuthorship W4386076265A5010479652 @default.
• W4386076265 hasAuthorship W4386076265A5019381882 @default.
• W4386076265 hasAuthorship W4386076265A5022256556 @default.
• W4386076265 hasAuthorship W4386076265A5024401174 @default.
• W4386076265 hasAuthorship W4386076265A5057135582 @default.
• W4386076265 hasAuthorship W4386076265A5064414647 @default.
• W4386076265 hasAuthorship W4386076265A5065035829 @default.
• W4386076265 hasConcept C154945302 @default.
• W4386076265 hasConcept C17744445 @default.
• W4386076265 hasConcept C185592680 @default.
• W4386076265 hasConcept C188027245 @default.
• W4386076265 hasConcept C199539241 @default.
• W4386076265 hasConcept C204321447 @default.
• W4386076265 hasConcept C23123220 @default.
• W4386076265 hasConcept C2776359362 @default.
• W4386076265 hasConcept C3018412434 @default.
• W4386076265 hasConcept C41008148 @default.
• W4386076265 hasConcept C49774154 @default.
• W4386076265 hasConcept C71139939 @default.
• W4386076265 hasConcept C94625758 @default.
• W4386076265 hasConceptScore W4386076265C154945302 @default.
• W4386076265 hasConceptScore W4386076265C17744445 @default.
• W4386076265 hasConceptScore W4386076265C185592680 @default.
• W4386076265 hasConceptScore W4386076265C188027245 @default.
• W4386076265 hasConceptScore W4386076265C199539241 @default.
• W4386076265 hasConceptScore W4386076265C204321447 @default.
• W4386076265 hasConceptScore W4386076265C23123220 @default.
• W4386076265 hasConceptScore W4386076265C2776359362 @default.
• W4386076265 hasConceptScore W4386076265C3018412434 @default.
• W4386076265 hasConceptScore W4386076265C41008148 @default.
• W4386076265 hasConceptScore W4386076265C49774154 @default.
• W4386076265 hasConceptScore W4386076265C71139939 @default.
• W4386076265 hasConceptScore W4386076265C94625758 @default.
• W4386076265 hasLocation W43860762651 @default.
• W4386076265 hasOpenAccess W4386076265 @default.
• W4386076265 hasPrimaryLocation W43860762651 @default.
• W4386076265 hasRelatedWork W2086064646 @default.
• W4386076265 hasRelatedWork W2115485936 @default.
• W4386076265 hasRelatedWork W2119135658 @default.
• W4386076265 hasRelatedWork W2293457016 @default.
• W4386076265 hasRelatedWork W2349174110 @default.
• W4386076265 hasRelatedWork W2357241418 @default.
• W4386076265 hasRelatedWork W2789919619 @default.
• W4386076265 hasRelatedWork W2792377126 @default.
• W4386076265 hasRelatedWork W3022131925 @default.
• W4386076265 hasRelatedWork W3169305685 @default.
• W4386076265 isParatext "false" @default.

• next