Many automated processes such as auto-piloting rely on a good semantic segmentation as a critical component. To speed up performance, it is common to downsample the input frame. However, this comes at the cost of missed small objects and reduced accuracy at semantic boundaries. To address this problem, we propose a new content-adaptive downsampling technique that learns to favor sampling locations near semantic boundaries of target classes. Costperformance analysis shows that our method consistently outperforms the uniform sampling improving balance between accuracy and computational efficiency. Our adaptive sampling gives segmentation with better quality of boundaries and more reliable support for smaller-size objects.

No abstract

No abstract

We propose a new segmentation model combining common regularization energies, e.g. Markov Random Field (MRF) potentials, and standard pairwise clustering criteria like Normalized Cut (NC), average association (AA), etc. These clustering and regularization models are widely used in machine learning and computer vision, but they were not combined before due to significant differences in the corresponding optimization, e.g. spectral relaxation and combinatorial max-flow techniques. On the one hand, we show that many common applications using MRF segmentation energies can benefit from a high-order NC term, e.g. enforcing balanced clustering of arbitrary high-dimensional image features combining color, texture, location, depth, motion, etc. On the other hand, standard clustering applications can benefit from an inclusion of common pairwise or higher-order MRF constraints, e.g. edge alignment, bin-consistency, label cost, etc. To address joint energies like NC+MRF, we propose efficient Kernel Cut algorithms based on bound optimization. While focusing on graph cut and move-making techniques, our new unary (linear) kernel and spectral bound formulations for common pairwise clustering criteria allow to integrate them with any regularization functionals with existing discrete or continuous solvers. arXiv:1506.07439v6 [cs.CV] 21 Sep 2016equivalent energy formulations: equivalent energy formulations:k related examples: related examples: elliptic K-means [31], [32] Average Association or Distortion [38] geometric model fitting [12] Average Cut [8] K-modes [29] or mean-shift [39] (weak kKM) Normalized Cut [8], [40] (weighted kKM) Entropy criterion ∑ k S k ⋅ H(S k ) [23], [26] Gini criterion ∑ k S k ⋅ G(S k ) [35], [41] for highly descriptive models (GMMs, histograms) for small-width normalized kernels (see Sec.5.1) Yuri Boykov received "Diploma of Higher Education" with honors at Moscow Institute of Physics and Technology (department of Radio Engineering and Cybernetics) in 1992 and completed his Ph.D. at the department of Operations Research at Cornell University in 1996. He is currently a full professor at the department of Computer Science at the University of Western Ontario. His research is concentrated in the area of computer vision and biomedical image analysis.In particular, he is interested in problems of early vision, image segmentation, restoration, registration, stereo, motion, model fitting, feature-based object recognition, photo-video editing and others. He is a recipient