Full-motion cost volumes play a central role in current state-of-the-art optical flow methods.
However, constructed using simple feature correlations, they lack the ability to encapsulate …

We introduce multigrid Predictive Filter Flow (mgPFF), a framework for unsupervised
learning on videos. The mgPFF takes as input a pair of frames and outputs per-pixel filters to …

Efficient optical flow estimation with high accuracy is a challenging problem in computer
vision. In this paper, we present a simple but efficient segmentation-based PatchMatch …

Constraints in the context of Machine Learning or Computer Vision play a central role to
obtain accurate models of the overall system or to simply prescribe its behavior under …

The impact of numerical optimization on modern data analysis has been quite significant.
Today, these methods lie at the heart of most statistical machine learning applications in …