Matches in SemOpenAlex for { <https://semopenalex.org/work/W3164831349> ?p ?o ?g. }
- W3164831349 endingPage "76285" @default.
- W3164831349 startingPage "76270" @default.
- W3164831349 abstract "Violence recognition is challenging since recognition must be performed on videos acquired by a lot of surveillance cameras at any time or place. It should make reliable detections in real time and inform surveillance personnel promptly when violent crimes take place. Therefore, we focus on efficient violence recognition for real-time and on-device operation, for easy expansion into a surveillance system with numerous cameras. In this paper, we propose a novel violence detection pipeline that can be combined with the conventional 2-dimensional Convolutional Neural Networks (2D CNNs). In particular, frame-grouping is proposed to give the 2D CNNs the ability to learn spatio-temporal representations in videos. It is a simple processing method to average the channels of input frames and group three consecutive channel-averaged frames as an input of the 2D CNNs. Furthermore, we present spatial and temporal attention modules that are lightweight but consistently improve the performance of violence recognition. The spatial attention module named Motion Saliency Map (MSM) can capture salient regions of feature maps derived from the motion boundaries using the difference between consecutive frames. The temporal attention module called Temporal Squeeze-and-Excitation (T-SE) block can inherently highlight the time periods that are correlated with a target event. Our proposed pipeline brings significant performance improvements compared to the 2D CNNs followed by the Long Short-Term Memory (LSTM) and much less computational complexity than existing 3D-CNN-based methods. In particular, MobileNetV3 and EfficientNet-B0 with our proposed modules achieved state-of-the-art performance on six different violence datasets. Our codes are available at https://github.com/ahstarwab/Violence_Detection." @default.
- W3164831349 created "2021-06-07" @default.
- W3164831349 creator A5027304168 @default.
- W3164831349 creator A5031280750 @default.
- W3164831349 creator A5045146037 @default.
- W3164831349 date "2021-01-01" @default.
- W3164831349 modified "2023-10-16" @default.
- W3164831349 title "Efficient Spatio-Temporal Modeling Methods for Real-Time Violence Recognition" @default.
- W3164831349 cites W1522734439 @default.
- W3164831349 cites W1902041153 @default.
- W3164831349 cites W1923404803 @default.
- W3164831349 cites W1947481528 @default.
- W3164831349 cites W1969914528 @default.
- W3164831349 cites W1990042487 @default.
- W3164831349 cites W2108598243 @default.
- W3164831349 cites W2129920697 @default.
- W3164831349 cites W2163588226 @default.
- W3164831349 cites W2233116163 @default.
- W3164831349 cites W2611015177 @default.
- W3164831349 cites W2738042021 @default.
- W3164831349 cites W2765644949 @default.
- W3164831349 cites W2883429621 @default.
- W3164831349 cites W2895738954 @default.
- W3164831349 cites W2940457086 @default.
- W3164831349 cites W2943934283 @default.
- W3164831349 cites W2951914530 @default.
- W3164831349 cites W2962735857 @default.
- W3164831349 cites W2962858109 @default.
- W3164831349 cites W2962943250 @default.
- W3164831349 cites W2963091558 @default.
- W3164831349 cites W2963145730 @default.
- W3164831349 cites W2963155035 @default.
- W3164831349 cites W2963163009 @default.
- W3164831349 cites W2963246338 @default.
- W3164831349 cites W2963420686 @default.
- W3164831349 cites W2963447094 @default.
- W3164831349 cites W2963524571 @default.
- W3164831349 cites W2963795951 @default.
- W3164831349 cites W2963820951 @default.
- W3164831349 cites W2963918968 @default.
- W3164831349 cites W2964217848 @default.
- W3164831349 cites W2972940264 @default.
- W3164831349 cites W2982083293 @default.
- W3164831349 cites W2990152177 @default.
- W3164831349 cites W2990525852 @default.
- W3164831349 cites W2990571558 @default.
- W3164831349 cites W2992457155 @default.
- W3164831349 cites W2999794487 @default.
- W3164831349 cites W3009112254 @default.
- W3164831349 cites W3011398234 @default.
- W3164831349 cites W3014564646 @default.
- W3164831349 cites W3036001641 @default.
- W3164831349 cites W3038667391 @default.
- W3164831349 cites W3089682612 @default.
- W3164831349 cites W3092481602 @default.
- W3164831349 cites W3093359244 @default.
- W3164831349 cites W3097283600 @default.
- W3164831349 cites W3106108873 @default.
- W3164831349 cites W3131207683 @default.
- W3164831349 doi "https://doi.org/10.1109/access.2021.3083273" @default.
- W3164831349 hasPublicationYear "2021" @default.
- W3164831349 type Work @default.
- W3164831349 sameAs 3164831349 @default.
- W3164831349 citedByCount "23" @default.
- W3164831349 countsByYear W31648313492021 @default.
- W3164831349 countsByYear W31648313492022 @default.
- W3164831349 countsByYear W31648313492023 @default.
- W3164831349 crossrefType "journal-article" @default.
- W3164831349 hasAuthorship W3164831349A5027304168 @default.
- W3164831349 hasAuthorship W3164831349A5031280750 @default.
- W3164831349 hasAuthorship W3164831349A5045146037 @default.
- W3164831349 hasBestOaLocation W31648313491 @default.
- W3164831349 hasConcept C104114177 @default.
- W3164831349 hasConcept C120665830 @default.
- W3164831349 hasConcept C121332964 @default.
- W3164831349 hasConcept C126042441 @default.
- W3164831349 hasConcept C138885662 @default.
- W3164831349 hasConcept C153180895 @default.
- W3164831349 hasConcept C154945302 @default.
- W3164831349 hasConcept C192209626 @default.
- W3164831349 hasConcept C199360897 @default.
- W3164831349 hasConcept C2524010 @default.
- W3164831349 hasConcept C2776401178 @default.
- W3164831349 hasConcept C2777210771 @default.
- W3164831349 hasConcept C2780719617 @default.
- W3164831349 hasConcept C31972630 @default.
- W3164831349 hasConcept C33923547 @default.
- W3164831349 hasConcept C41008148 @default.
- W3164831349 hasConcept C41895202 @default.
- W3164831349 hasConcept C43521106 @default.
- W3164831349 hasConcept C76155785 @default.
- W3164831349 hasConcept C81363708 @default.
- W3164831349 hasConceptScore W3164831349C104114177 @default.
- W3164831349 hasConceptScore W3164831349C120665830 @default.
- W3164831349 hasConceptScore W3164831349C121332964 @default.
- W3164831349 hasConceptScore W3164831349C126042441 @default.
- W3164831349 hasConceptScore W3164831349C138885662 @default.
- W3164831349 hasConceptScore W3164831349C153180895 @default.