NII, Japan at MediaEval 2012 Violent Scenes Detection Affect Task

Violent scene detection (VSD) is a challenging problem because of the heterogeneous content, large variations in video quality, and semantic meaning of the concepts. The Violent Scenes Detection Task of MediaEval [1] provides a common dataset and evaluation protocol thus enables a fair comparison of methods. In this paper, we describe our VSD system used in MediaEval 2014 and briey discuss the performance results obtained in main subjective tasks. In this year, we focus on improving the trajectory-based motion features that have been proven eective in previous year’s evaluation. Besides that, we also adopt SIFT-based and audio features as in last year’s system. We combined these features using late fusion. Our results show that the trajectory-based motion features still have very competitive performance and the combination with still image features and audio features can improve overall performance.

[1]  Mohammad Soleymani,et al.  The MediaEval 2011 Affect Task: Violent Scenes Detection in Hollywood movies , 2011, MediaEval.

[2]  Koen E. A. van de Sande,et al.  Evaluating Color Descriptors for Object and Scene Recognition , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[4]  Cordelia Schmid,et al.  Action Recognition with Improved Trajectories , 2013, 2013 IEEE International Conference on Computer Vision.