Human Attention Model for Action Movie Analysis

Nowadays, automatic detection of highlights for movies is indispensable for video management and browsing. In this paper, we specifically present the formulation of human attention model and its application for action movie analysis. Depending on the relationship between stimuli in both visual and audio modalities and the change of human attention, we construct the visual and audio attention sub-models respectively. By integrating both sub-models, human attention model is formulated and attention flow plot is derived to simulate the change of human attention in the time domain. Based on video structuralization and human attention analysis, we realize action concept annotation for interest-oriented navigation of viewers. The large-scale experiments demonstrate the effectiveness and generality of human attention model for action movie analysis.

[1]  Svetha Venkatesh,et al.  Novel approach to determining tempo and dramatic story sections in motion pictures , 2000, Proceedings 2000 International Conference on Image Processing (Cat. No.00CH37101).

[2]  Christof Koch,et al.  A Model of Saliency-Based Visual Attention for Rapid Scene Analysis , 2009 .

[3]  Alan F. Smeaton,et al.  Automatically selecting shots for action movie trailers , 2006, MIR '06.

[4]  Svetha Venkatesh,et al.  Toward automatic extraction of expressive elements from motion pictures: tempo , 2002, IEEE Trans. Multim..

[5]  Sheng Tang,et al.  TRECVID 2006 Rushes Exploitation by CAS MCG , 2006, TRECVID.

[6]  Lei Chen,et al.  Incorporating Audio Cues into Dialog and Action Scene Extraction , 2003, IS&T/SPIE Electronic Imaging.

[7]  Bai Liang,et al.  Feature analysis and extraction for audio automatic classification , 2005, 2005 IEEE International Conference on Systems, Man and Cybernetics.

[8]  Wei-Ta Chu,et al.  Action movies segmentation and summarization based on tempo analysis , 2004, MIR '04.

[9]  C. Koch,et al.  Computational modelling of visual attention , 2001, Nature Reviews Neuroscience.