文献类型：期刊文献

英文题名：Context-Aware Mouse Behavior Recognition Using Hidden Markov Models

作者：Jiang, Zheheng[1];Crookes, Danny[2];Green, Brian D.[3];Zhao, Yunfeng[2];Ma, Haiping[4];Li, Ling[5];Zhang, Shengping[6];Tao, Dacheng[7];Zhou, Huiyu[1]

机构：[1]Univ Leicester, Dept Informat, Leicester LE1 7RH, Leics, England;[2]Queens Univ Belfast, Sch Elect Elect Engn & Comp Sci, Belfast BT3 9DT, Antrim, North Ireland;[3]Queens Univ Belfast, Sch Biol Sci, Belfast BT3 9DT, Antrim, North Ireland;[4]Shaoxing Univ, Dept Elect Engn, Shaoxing 312000, Peoples R China;[5]Univ Kent, Sch Comp, Canterbury CT2 7NZ, Kent, England;[6]Harbin Inst Technol, Sch Comp Sci & Technol, Weihai 264200, Peoples R China;[7]Univ Sydney, UBTECH Sydney Artificial Intelligence Ctr, Fac Engn & Informat Technol, Sch Informat Technol, Darlington, NSW 2008, Australia

年份：2019

卷号：28

期号：3

起止页码：1133

外文期刊名：IEEE TRANSACTIONS ON IMAGE PROCESSING

收录：SCI-EXPANDED(收录号：WOS:000449398900001)、、EI(收录号：20184205949127)、Scopus(收录号：2-s2.0-85054611431)、WOS

基金：This work was supported by the U.K. EPSRC under Grant EP/N011074/1. The work of H. Ma was supported by the Zhejiang Provincial Natural Science Foundation of China under Grant Y19F030029. The work of S. Zhang was supported by the National Natural Science Foundation of China under Grant 61872112. The work of D. Tao was supported by the Australian Research Council Projects under Grants FL-170100117, DP-180103424, and IH180100002. The work of H. Zhou was supported by the Royal Society-Newton Advanced Fellowship under Grant NA160342.

语种：英文

外文关键词：Mouse behaviors; hidden Markov model; spatial-temporal segment; Fisher vector; segment aggregate network

外文摘要：Automated recognition of mouse behaviors is crucial in studying psychiatric and neurologic diseases. To achieve this objective, it is very important to analyze the temporal dynamics of mouse behaviors. In particular, the change between mouse neighboring actions is swift in a short period. In this paper, we develop and implement a novel hidden Markov model (HMM) algorithm to describe the temporal characteristics of mouse behaviors. In particular, we here propose a hybrid deep learning architecture, where the first unsupervised layer relies on an advanced spatial-temporal segment Fisher vector encoding both visual and contextual features. Subsequent supervised layers based on our segment aggregate network are trained to estimate the state-dependent observation probabilities of the HMM. The proposed architecture shows the ability to discriminate between visually similar behaviors and results in high recognition rates with the strength of processing imbalanced mouse behavior datasets. Finally, we evaluate our approach using JHuang's and our own datasets, and the results show that our method outperforms other state-of-the-art approaches.