Matches in SemOpenAlex for { <https://semopenalex.org/work/W2006386748> ?p ?o ?g. }
- W2006386748 abstract "In this paper, we propose an unsupervised framework for action spotting in videos that does not depend on any specific feature (e.g. HOG/HOF, STIP, silhouette, bag-of-words, etc.). Furthermore, our solution requires no human localization, segmentation, or framewise tracking. This is achieved by treating the problem holistically as that of extracting the internal dynamics of video cuboids by modeling them in their natural form as multilinear tensors. To extract their internal dynamics, we devised a novel Two-Phase Decomposition (TP-Decomp) of a tensor that generates very compact and discriminative representations that are robust to even heavily perturbed data. Technically, a Rank-based Tensor Core Pyramid (Rank-TCP) descriptor is generated by combining multiple tensor cores under multiple ranks, allowing to represent video cuboids in a hierarchical tensor pyramid. The problem then reduces to a template matching problem, which is solved efficiently by using two boosting strategies: (1) to reduce search space, we filter the dense trajectory cloud extracted from the target video, (2) to boost the matching speed, we perform matching in an iterative coarse-to-fine manner. Experiments on 5 benchmarks show that our method outperforms current state-of-the-art under various challenging conditions. We also created a challenging dataset called Heavily Perturbed Video Array (HPVA) to validate the robustness of our framework under heavily perturbed situations." @default.
- W2006386748 created "2016-06-24" @default.
- W2006386748 creator A5058575676 @default.
- W2006386748 creator A5073612283 @default.
- W2006386748 creator A5076117344 @default.
- W2006386748 date "2014-06-01" @default.
- W2006386748 modified "2023-09-27" @default.
- W2006386748 title "Feature-Independent Action Spotting without Human Localization, Segmentation, or Frame-wise Tracking" @default.
- W2006386748 cites W137381930 @default.
- W2006386748 cites W1482825550 @default.
- W2006386748 cites W1979606776 @default.
- W2006386748 cites W1996904744 @default.
- W2006386748 cites W2024165284 @default.
- W2006386748 cites W2024166170 @default.
- W2006386748 cites W2041941194 @default.
- W2006386748 cites W2051849793 @default.
- W2006386748 cites W2054674781 @default.
- W2006386748 cites W2067766814 @default.
- W2006386748 cites W2068611653 @default.
- W2006386748 cites W2097947185 @default.
- W2006386748 cites W2106221905 @default.
- W2006386748 cites W2108871361 @default.
- W2006386748 cites W2111448058 @default.
- W2006386748 cites W2113055885 @default.
- W2006386748 cites W2113201641 @default.
- W2006386748 cites W2113800095 @default.
- W2006386748 cites W2116931983 @default.
- W2006386748 cites W2122319321 @default.
- W2006386748 cites W2124017786 @default.
- W2006386748 cites W2135658380 @default.
- W2006386748 cites W2137981002 @default.
- W2006386748 cites W2149126415 @default.
- W2006386748 cites W2162451865 @default.
- W2006386748 cites W2162659160 @default.
- W2006386748 cites W2172207578 @default.
- W2006386748 cites W2533739470 @default.
- W2006386748 cites W2539525772 @default.
- W2006386748 cites W2547599103 @default.
- W2006386748 cites W67471658 @default.
- W2006386748 doi "https://doi.org/10.1109/cvpr.2014.344" @default.
- W2006386748 hasPublicationYear "2014" @default.
- W2006386748 type Work @default.
- W2006386748 sameAs 2006386748 @default.
- W2006386748 citedByCount "17" @default.
- W2006386748 countsByYear W20063867482015 @default.
- W2006386748 countsByYear W20063867482016 @default.
- W2006386748 countsByYear W20063867482017 @default.
- W2006386748 countsByYear W20063867482018 @default.
- W2006386748 countsByYear W20063867482020 @default.
- W2006386748 crossrefType "proceedings-article" @default.
- W2006386748 hasAuthorship W2006386748A5058575676 @default.
- W2006386748 hasAuthorship W2006386748A5073612283 @default.
- W2006386748 hasAuthorship W2006386748A5076117344 @default.
- W2006386748 hasConcept C104317684 @default.
- W2006386748 hasConcept C153180895 @default.
- W2006386748 hasConcept C154945302 @default.
- W2006386748 hasConcept C155281189 @default.
- W2006386748 hasConcept C185592680 @default.
- W2006386748 hasConcept C202444582 @default.
- W2006386748 hasConcept C2986737658 @default.
- W2006386748 hasConcept C31972630 @default.
- W2006386748 hasConcept C33923547 @default.
- W2006386748 hasConcept C41008148 @default.
- W2006386748 hasConcept C42704193 @default.
- W2006386748 hasConcept C52622490 @default.
- W2006386748 hasConcept C55493867 @default.
- W2006386748 hasConcept C63479239 @default.
- W2006386748 hasConcept C89600930 @default.
- W2006386748 hasConcept C97931131 @default.
- W2006386748 hasConceptScore W2006386748C104317684 @default.
- W2006386748 hasConceptScore W2006386748C153180895 @default.
- W2006386748 hasConceptScore W2006386748C154945302 @default.
- W2006386748 hasConceptScore W2006386748C155281189 @default.
- W2006386748 hasConceptScore W2006386748C185592680 @default.
- W2006386748 hasConceptScore W2006386748C202444582 @default.
- W2006386748 hasConceptScore W2006386748C2986737658 @default.
- W2006386748 hasConceptScore W2006386748C31972630 @default.
- W2006386748 hasConceptScore W2006386748C33923547 @default.
- W2006386748 hasConceptScore W2006386748C41008148 @default.
- W2006386748 hasConceptScore W2006386748C42704193 @default.
- W2006386748 hasConceptScore W2006386748C52622490 @default.
- W2006386748 hasConceptScore W2006386748C55493867 @default.
- W2006386748 hasConceptScore W2006386748C63479239 @default.
- W2006386748 hasConceptScore W2006386748C89600930 @default.
- W2006386748 hasConceptScore W2006386748C97931131 @default.
- W2006386748 hasLocation W20063867481 @default.
- W2006386748 hasOpenAccess W2006386748 @default.
- W2006386748 hasPrimaryLocation W20063867481 @default.
- W2006386748 hasRelatedWork W1954643962 @default.
- W2006386748 hasRelatedWork W1973271201 @default.
- W2006386748 hasRelatedWork W1973715457 @default.
- W2006386748 hasRelatedWork W1974598519 @default.
- W2006386748 hasRelatedWork W1975088888 @default.
- W2006386748 hasRelatedWork W2018805841 @default.
- W2006386748 hasRelatedWork W2051849793 @default.
- W2006386748 hasRelatedWork W2074763219 @default.
- W2006386748 hasRelatedWork W2086441628 @default.
- W2006386748 hasRelatedWork W2092920974 @default.
- W2006386748 hasRelatedWork W2106271850 @default.
- W2006386748 hasRelatedWork W2153719684 @default.