Matches in SemOpenAlex for { <https://semopenalex.org/work/W4225375635> ?p ?o ?g. }
Showing items 1 to 51 of
51
with 100 items per page.
- W4225375635 abstract "In weakly-supervised temporal action localization (WS-TAL), the methods commonly follow the localization by classification procedure, which uses the snippet predictions to form video class scores and then optimizes a video classification loss. In this procedure, the snippet predictions (or snippet attention weights) are used to separate foreground and background. However, the snippet predictions are usually inaccurate due to absence of frame-wise labels, and then the overall performance is hindered. In this paper, we propose a novel C$^3$BN to achieve robust snippet predictions. C$^3$BN includes two key designs by exploring the inherent characteristics of video data. First, because of the natural continuity of adjacent snippets, we propose a micro data augmentation strategy to increase the diversity of snippets with convex combination of adjacent snippets. Second, we propose a macro-micro consistency regularization strategy to force the model to be invariant (or equivariant) to the transformations of snippets with respect to video semantics, snippet predictions and snippet features. Experimental results demonstrate the effectiveness of our proposed method on top of baselines for the WS-TAL tasks with video-level and point-level supervision." @default.
- W4225375635 created "2022-05-05" @default.
- W4225375635 creator A5036125002 @default.
- W4225375635 creator A5065292415 @default.
- W4225375635 creator A5077703661 @default.
- W4225375635 creator A5079853816 @default.
- W4225375635 date "2022-05-01" @default.
- W4225375635 modified "2023-09-24" @default.
- W4225375635 title "Convex Combination Consistency between Neighbors for Weakly-supervised Action Localization" @default.
- W4225375635 doi "https://doi.org/10.48550/arxiv.2205.00400" @default.
- W4225375635 hasPublicationYear "2022" @default.
- W4225375635 type Work @default.
- W4225375635 citedByCount "0" @default.
- W4225375635 crossrefType "posted-content" @default.
- W4225375635 hasAuthorship W4225375635A5036125002 @default.
- W4225375635 hasAuthorship W4225375635A5065292415 @default.
- W4225375635 hasAuthorship W4225375635A5077703661 @default.
- W4225375635 hasAuthorship W4225375635A5079853816 @default.
- W4225375635 hasBestOaLocation W42253756351 @default.
- W4225375635 hasConcept C119857082 @default.
- W4225375635 hasConcept C154945302 @default.
- W4225375635 hasConcept C23123220 @default.
- W4225375635 hasConcept C2776135515 @default.
- W4225375635 hasConcept C2776436953 @default.
- W4225375635 hasConcept C2777822670 @default.
- W4225375635 hasConcept C41008148 @default.
- W4225375635 hasConcept C5274069 @default.
- W4225375635 hasConceptScore W4225375635C119857082 @default.
- W4225375635 hasConceptScore W4225375635C154945302 @default.
- W4225375635 hasConceptScore W4225375635C23123220 @default.
- W4225375635 hasConceptScore W4225375635C2776135515 @default.
- W4225375635 hasConceptScore W4225375635C2776436953 @default.
- W4225375635 hasConceptScore W4225375635C2777822670 @default.
- W4225375635 hasConceptScore W4225375635C41008148 @default.
- W4225375635 hasConceptScore W4225375635C5274069 @default.
- W4225375635 hasLocation W42253756351 @default.
- W4225375635 hasOpenAccess W4225375635 @default.
- W4225375635 hasPrimaryLocation W42253756351 @default.
- W4225375635 hasRelatedWork W2350879319 @default.
- W4225375635 hasRelatedWork W2961085424 @default.
- W4225375635 hasRelatedWork W3046775127 @default.
- W4225375635 hasRelatedWork W4220812971 @default.
- W4225375635 hasRelatedWork W4285260836 @default.
- W4225375635 hasRelatedWork W4286629047 @default.
- W4225375635 hasRelatedWork W4295135962 @default.
- W4225375635 hasRelatedWork W4306321456 @default.
- W4225375635 hasRelatedWork W4306674287 @default.
- W4225375635 hasRelatedWork W4224009465 @default.
- W4225375635 isParatext "false" @default.
- W4225375635 isRetracted "false" @default.
- W4225375635 workType "article" @default.