70 likes | 237 Vues
Exemplar-SVM for Action Recognition. Week 10 Presented by Christina Peterson. Movement Exemplar-SVMs . Tran and Torresani [1] based the MEX-SVM on the work of Malisiewicz et. al. [2]
E N D
Exemplar-SVM for Action Recognition Week 10 Presented by Christina Peterson
Movement Exemplar-SVMs • Tran and Torresani [1] based the MEX-SVM on the work of Malisiewiczet. al. [2] • Linear SVMs applied to histograms of space-time interest points (STIPs) calculated from sub-volumes of the video • Trained on one positive samples and many negative samples • Calibrate MEX-SVM’s using Platt’s Method
Reasons for Discrepancies • Different training/testing set • MEX-SVMs trained on UCF50 data set, tested on HMDB51 • Exemplar-SVM trained and tested on UCF50 data set • Exemplar Feature Vector • MEX-SVM used ground truth bounding box • Exemplar-SVM use entire video • Mid-Level Feature Vector • MEX-SVM Mid-Level Feature Dimension = Na x Ns x Np • Na = Number of Exemplars • Ns = Exemplar template scale • Np = Spatial-Temporal Pyramid Level • 185 x 3 x (1 + 8 + 64) = 40,515 • Exemplar-SVM Mid-Level Feature Dimension = Na • Varied between 250 – 1,500
References [1] D. Tran and L. Torresani. MEXSVMs: Mid-level Features for Scalable Action Recognition. Dartmouth Computer Science Techinical Report TR2013-726, January 2013. [2] T. Malisiewicz, A. Gupta, and A. A. Efros. Ensemble of Exemplar SVMS for Object Detection and Beyond. In Proc. ICCV, 2011.