报告题目： Automated recognition of mouse behaviours
报告人: Huiyu (Joe) Zhou, Reader,（英国University of Leicester）
Automated recognition of mouse behaviours is crucial in studying psychiatric and neurologic diseases, e.g. Parkinson’s disease. To achieve this objective, it is very important to analyse temporal dynamics of mouse behaviours. In this paper, we develop and implement a novel Hidden Markov Model (HMM) algorithm to describe the temporal characteristics of mouse behaviours. In particular, we here propose a hybrid deep learning architecture, where the first unsupervised layer relies on a new spatial-temporal segment Fisher Vector (SFV) encoding both visual and contextual features. Subsequent supervised layers based on our segment aggregated network (SAN) are trained to estimate the state dependent observation probabilities of the HMM. 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.