Matches in SemOpenAlex for { <https://semopenalex.org/work/W3208474254> ?p ?o ?g. }
- W3208474254 abstract "Due to the large memory footprint of untrimmed videos, current state-of-the-art video localization methods operate atop precomputed video clip features. These features are extracted from video encoders typically trained for trimmed action classification tasks, making such features not necessarily suitable for temporal localization. In this work, we propose a novel supervised pretraining paradigm for clip features that not only trains to classify activities but also considers background clips and global video information to improve temporal sensitivity. Extensive experiments show that using features trained with our novel pretraining strategy significantly improves the performance of recent state-of-the-art methods on three tasks: Temporal Action Localization, Action Proposal Generation, and Dense Video Captioning. We also show that our pretraining approach is effective across three encoder architectures and two pretraining datasets. We believe video feature encoding is an important building block for localization algorithms, and extracting temporally-sensitive features should be of paramount importance in building more accurate models. The code and pretrained models are available on our project website." @default.
- W3208474254 created "2021-11-08" @default.
- W3208474254 creator A5024763828 @default.
- W3208474254 creator A5055514112 @default.
- W3208474254 creator A5091475013 @default.
- W3208474254 date "2021-10-01" @default.
- W3208474254 modified "2023-10-05" @default.
- W3208474254 title "TSP: Temporally-Sensitive Pretraining of Video Encoders for Localization Tasks" @default.
- W3208474254 cites W1522734439 @default.
- W3208474254 cites W1927052826 @default.
- W3208474254 cites W2009139688 @default.
- W3208474254 cites W2016053056 @default.
- W3208474254 cites W2084341401 @default.
- W3208474254 cites W2194775991 @default.
- W3208474254 cites W2302086703 @default.
- W3208474254 cites W2336403884 @default.
- W3208474254 cites W2507009361 @default.
- W3208474254 cites W2519328139 @default.
- W3208474254 cites W2575842049 @default.
- W3208474254 cites W2593116425 @default.
- W3208474254 cites W2593722617 @default.
- W3208474254 cites W2597958930 @default.
- W3208474254 cites W2745461083 @default.
- W3208474254 cites W2755876276 @default.
- W3208474254 cites W2883910824 @default.
- W3208474254 cites W2884293275 @default.
- W3208474254 cites W2884969173 @default.
- W3208474254 cites W2895240652 @default.
- W3208474254 cites W2948242301 @default.
- W3208474254 cites W2951098185 @default.
- W3208474254 cites W2952435096 @default.
- W3208474254 cites W2955874753 @default.
- W3208474254 cites W2962799512 @default.
- W3208474254 cites W2962876901 @default.
- W3208474254 cites W2962990649 @default.
- W3208474254 cites W2963037989 @default.
- W3208474254 cites W2963045696 @default.
- W3208474254 cites W2963150697 @default.
- W3208474254 cites W2963155035 @default.
- W3208474254 cites W2963177403 @default.
- W3208474254 cites W2963247196 @default.
- W3208474254 cites W2963321993 @default.
- W3208474254 cites W2963351113 @default.
- W3208474254 cites W2963524571 @default.
- W3208474254 cites W2963631366 @default.
- W3208474254 cites W2963753226 @default.
- W3208474254 cites W2963874033 @default.
- W3208474254 cites W2963916161 @default.
- W3208474254 cites W2964008341 @default.
- W3208474254 cites W2964214371 @default.
- W3208474254 cites W2964216549 @default.
- W3208474254 cites W2964274041 @default.
- W3208474254 cites W2968104955 @default.
- W3208474254 cites W2980037812 @default.
- W3208474254 cites W2982083293 @default.
- W3208474254 cites W2983918066 @default.
- W3208474254 cites W2984287396 @default.
- W3208474254 cites W2984478308 @default.
- W3208474254 cites W2984862483 @default.
- W3208474254 cites W2985144848 @default.
- W3208474254 cites W2986407524 @default.
- W3208474254 cites W2988753485 @default.
- W3208474254 cites W2990152177 @default.
- W3208474254 cites W2990503944 @default.
- W3208474254 cites W2997314266 @default.
- W3208474254 cites W2997706915 @default.
- W3208474254 cites W2998582438 @default.
- W3208474254 cites W2998702159 @default.
- W3208474254 cites W3004505825 @default.
- W3208474254 cites W3034221024 @default.
- W3208474254 cites W3034623254 @default.
- W3208474254 cites W3034971973 @default.
- W3208474254 cites W3035251589 @default.
- W3208474254 cites W3035392611 @default.
- W3208474254 cites W3092820804 @default.
- W3208474254 cites W3095669214 @default.
- W3208474254 cites W3106041614 @default.
- W3208474254 cites W3110589170 @default.
- W3208474254 cites W3118479201 @default.
- W3208474254 doi "https://doi.org/10.1109/iccvw54120.2021.00356" @default.
- W3208474254 hasPublicationYear "2021" @default.
- W3208474254 type Work @default.
- W3208474254 sameAs 3208474254 @default.
- W3208474254 citedByCount "32" @default.
- W3208474254 countsByYear W32084742542022 @default.
- W3208474254 countsByYear W32084742542023 @default.
- W3208474254 crossrefType "proceedings-article" @default.
- W3208474254 hasAuthorship W3208474254A5024763828 @default.
- W3208474254 hasAuthorship W3208474254A5055514112 @default.
- W3208474254 hasAuthorship W3208474254A5091475013 @default.
- W3208474254 hasBestOaLocation W32084742542 @default.
- W3208474254 hasConcept C111919701 @default.
- W3208474254 hasConcept C115961682 @default.
- W3208474254 hasConcept C118505674 @default.
- W3208474254 hasConcept C125411270 @default.
- W3208474254 hasConcept C138885662 @default.
- W3208474254 hasConcept C153180895 @default.
- W3208474254 hasConcept C154945302 @default.
- W3208474254 hasConcept C157657479 @default.
- W3208474254 hasConcept C177264268 @default.