Matches in SemOpenAlex for { <https://semopenalex.org/work/W3166893724> ?p ?o ?g. }
- W3166893724 abstract "Most existing video-and-language (VidL) research focuses on a single dataset, or multiple datasets of a single task. In reality, a truly useful VidL system is expected to be easily generalizable to diverse tasks, domains, and datasets. To facilitate the evaluation of such systems, we introduce Video-And-Language Understanding Evaluation (VALUE) benchmark, an assemblage of 11 VidL datasets over 3 popular tasks: (i) text-to-video retrieval; (ii) video question answering; and (iii) video captioning. VALUE benchmark aims to cover a broad range of video genres, video lengths, data volumes, and task difficulty levels. Rather than focusing on single-channel videos with visual information only, VALUE promotes models that leverage information from both video frames and their associated subtitles, as well as models that share knowledge across multiple tasks. We evaluate various baseline methods with and without large-scale VidL pre-training, and systematically investigate the impact of video input channels, fusion methods, and different video representations. We also study the transferability between tasks, and conduct multi-task learning under different settings. The significant gap between our best model and human performance calls for future study for advanced VidL models. VALUE is available at this https URL." @default.
- W3166893724 created "2021-06-22" @default.
- W3166893724 creator A5001425662 @default.
- W3166893724 creator A5001987532 @default.
- W3166893724 creator A5007285444 @default.
- W3166893724 creator A5008309880 @default.
- W3166893724 creator A5012449118 @default.
- W3166893724 creator A5026746295 @default.
- W3166893724 creator A5028783832 @default.
- W3166893724 creator A5036418431 @default.
- W3166893724 creator A5037467245 @default.
- W3166893724 creator A5048295582 @default.
- W3166893724 creator A5050195037 @default.
- W3166893724 creator A5066666034 @default.
- W3166893724 creator A5077322975 @default.
- W3166893724 creator A5084879213 @default.
- W3166893724 creator A5091607703 @default.
- W3166893724 date "2021-06-08" @default.
- W3166893724 modified "2023-10-04" @default.
- W3166893724 title "VALUE: A Multi-Task Benchmark for Video-and-Language Understanding Evaluation" @default.
- W3166893724 cites W1840435438 @default.
- W3166893724 cites W1956340063 @default.
- W3166893724 cites W2078238240 @default.
- W3166893724 cites W2101105183 @default.
- W3166893724 cites W2108598243 @default.
- W3166893724 cites W2126400076 @default.
- W3166893724 cites W2133458109 @default.
- W3166893724 cites W2133459682 @default.
- W3166893724 cites W2154652894 @default.
- W3166893724 cites W2156303437 @default.
- W3166893724 cites W2163455955 @default.
- W3166893724 cites W2164290393 @default.
- W3166893724 cites W2194775991 @default.
- W3166893724 cites W2251861449 @default.
- W3166893724 cites W2251939518 @default.
- W3166893724 cites W2425121537 @default.
- W3166893724 cites W2462305634 @default.
- W3166893724 cites W2525778437 @default.
- W3166893724 cites W2606982687 @default.
- W3166893724 cites W2619947201 @default.
- W3166893724 cites W2765716052 @default.
- W3166893724 cites W2784025607 @default.
- W3166893724 cites W2806311723 @default.
- W3166893724 cites W2883429621 @default.
- W3166893724 cites W2899771611 @default.
- W3166893724 cites W2908510526 @default.
- W3166893724 cites W2963017553 @default.
- W3166893724 cites W2963310665 @default.
- W3166893724 cites W2963341956 @default.
- W3166893724 cites W2963351113 @default.
- W3166893724 cites W2963403868 @default.
- W3166893724 cites W2963541336 @default.
- W3166893724 cites W2963916161 @default.
- W3166893724 cites W2964089981 @default.
- W3166893724 cites W2964165804 @default.
- W3166893724 cites W2964345792 @default.
- W3166893724 cites W2965373594 @default.
- W3166893724 cites W2970597249 @default.
- W3166893724 cites W2979826702 @default.
- W3166893724 cites W2981851019 @default.
- W3166893724 cites W2984008963 @default.
- W3166893724 cites W2990503944 @default.
- W3166893724 cites W2990704537 @default.
- W3166893724 cites W2997805943 @default.
- W3166893724 cites W3023441976 @default.
- W3166893724 cites W3034188691 @default.
- W3166893724 cites W3034636873 @default.
- W3166893724 cites W3034727271 @default.
- W3166893724 cites W3034730770 @default.
- W3166893724 cites W3035276082 @default.
- W3166893724 cites W3035579820 @default.
- W3166893724 cites W3035635319 @default.
- W3166893724 cites W3043840704 @default.
- W3166893724 cites W3046692080 @default.
- W3166893724 cites W3102483398 @default.
- W3166893724 cites W3104862079 @default.
- W3166893724 cites W3105232955 @default.
- W3166893724 cites W3118106810 @default.
- W3166893724 cites W3118942129 @default.
- W3166893724 cites W3119786062 @default.
- W3166893724 cites W3122640483 @default.
- W3166893724 cites W3135367836 @default.
- W3166893724 cites W3139243190 @default.
- W3166893724 cites W3145235298 @default.
- W3166893724 cites W3152798676 @default.
- W3166893724 cites W3167366812 @default.
- W3166893724 cites W3169483174 @default.
- W3166893724 cites W3173449436 @default.
- W3166893724 cites W3175148590 @default.
- W3166893724 cites W3176362845 @default.
- W3166893724 cites W3204588463 @default.
- W3166893724 hasPublicationYear "2021" @default.
- W3166893724 type Work @default.
- W3166893724 sameAs 3166893724 @default.
- W3166893724 citedByCount "2" @default.
- W3166893724 countsByYear W31668937242021 @default.
- W3166893724 crossrefType "posted-content" @default.
- W3166893724 hasAuthorship W3166893724A5001425662 @default.
- W3166893724 hasAuthorship W3166893724A5001987532 @default.
- W3166893724 hasAuthorship W3166893724A5007285444 @default.