No abstract

This paper addresses the problem of silhouettebased human action segmentation and recognition in monocular sequences. Motion History Images (MHIs), used as 2D templates, capture motion information by encoding where and when motion occurred in the images. Inspired by codebook approaches for object and scene categorization, we first construct a codebook of temporal motion templates by clustering all the MHIs of each particular action. These MHIs capture different actors, speeds and a wide range of camera viewpoints. In this paper, we use a Kohonen's Self-Organizing Map (SOM) to simultaneously cluster the MHI templates and represent them in lower dimensional subspaces. To cope with temporal segmentation, and concurrently carry out action recognition, a new architecture is proposed where the obsrvation MHIs are projected onto all these actionspecific manifolds and the Euclidean distance between each MHI and the nearest cluster within each action-manifold constitutes the observation vector of a Markov Model. To estimate the state/action at each time step, we introduce a new method based on Observable Markov Models (OMMs) where the Markov model is augmented with a neutral state. The combination of our action-specific manifolds with the augmented OMM allows to automatically segment and recognize long sequences of consecutive actions, without any prior knowledge about initial and ending frames of each action. Importantly, our method allows to interpolate betweeen training viewpoint and recognizes actions, independently of the camera viewpoint, even from unseen viewpoints.

Developing a tool which identifies emotions based on their effect on cardiac activity may have a potential impact on clinical practice, since it may help in the diagnosing of psychoneural illnesses. In this study, a method based on the analysis of heart rate variability (HRV) guided by respiration is proposed. The method was based on redefining the high frequency (HF) band, not only to be centered at the respiratory frequency, but also to have a bandwidth dependent on the respiratory spectrum. The method was first tested using simulated HRV signals, yielding the minimum estimation errors as compared to classic and respiratory frequency centered at HF band based definitions, independently of the values of the sympathovagal ratio. Then, the proposed method was applied to discriminate emotions in a database of video-induced elicitation. Five emotional states, relax, joy, fear, sadness and anger, were considered. The maximum correlation between HRV and respiration spectra discriminated joy vs. relax, joy vs. each negative valence emotion, and fear vs. sadness with p-value ≤ 0.05 and AUC ≥ 0.70. Based on these results, human emotion characterization may be improved by adding respiratory information to HRV analysis.

No abstract

No abstract