Dynamic Attribute Package: Crowd Behavior Recognition in Complex Scene

Autor:	Lin Chen, Ji Zhu, Hua Yang, Tianqi Shi
Rok vydání:	2019
Předmět:	Curl (mathematics) Computer science business.industry Crowd management ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Behavior recognition Convolutional neural network Social dynamics Discriminative model Motion vector field Computer vision Artificial intelligence Crowd psychology business
Zdroj:	Digital TV and Multimedia Communication ISBN: 9789811381379 IFTC
DOI:	10.1007/978-981-13-8138-6_36
Popis:	Crowd behavior recognition under complex surveillance scenarios is a fundamental and important problem in crowd management application. In this paper, a comprehensive and specific overall-level dynamic attribute package is proposed by considering local pattern-related motion and group-level motion together to represent crowd movement. Curl and divergence map of normalized average motion vector field act as local pattern-related motion, which represents physical movement tendency of each particle. Group-level motion explores crowd interaction of inter-/intra-group, which focus on depicting crowd’s social dynamic property. The complementary characteristic of two motion representation in different level is analyzed and verified. Single frames in video clips and the corresponding dynamic attribute packages are sent into two-branch structured ConvNet, which can extract more discriminative spatial-temporal feature for behavior recognition. Experiment results conducted on CUHK dataset show that the proposed crowd behavior recognition framework outperforms than existing approaches and obtains the state-of-art performance.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::d7690e2e9c2d26150331faba838768a3 https://doi.org/10.1007/978-981-13-8138-6_36 Zobrazit plný text záznamu