Exploring Weak Stabilization for Motion Feature Extraction

Dennis Park, C. Lawrence Zitnick, Deva Ramanan, and Piotr Dollár

Abstract

We describe novel but simple motion features for the problem of detecting objects in video sequences. Previous approaches either compute optical flow or temporal differences on video frame pairs with various assumptions about stabilization. We describe a combined approach that uses coarse-scale flow and fine-scale temporal difference features. Our approach performs weak motion stabilization by factoring out camera motion and coarse object motion while preserving nonrigid motions that serve as useful cues for recognition. We show results for pedestrian detection and human pose estimation in video sequences, achieving state-of-the-art results in both. In particular, given a fixed detection rate our method achieves a five-fold reduction in false positives over prior art on the Caltech Pedestrian benchmark. Finally, we perform extensive diagnostic experiments to reveal what aspects of our system are crucial for good performance. Proper stabilization, long time-scale features, and proper normalization are all critical.

Details

Publication typeInproceedings
Published inCVPR
URLhttp://vision.ucsd.edu/~pdollar/research.html
PublisherComputer Vision and Pattern Recognition
> Publications > Exploring Weak Stabilization for Motion Feature Extraction