- Decision: conferenceOral
- Abstract: We present an inference procedure for the semantic segmentation of images %Different from many CRF approaches that are mostly %based on local potentials,%rely on dependencies modeled with unary and pairwise pixel or superpixel potentials, , based on estimates of the overlap between each of a set of large object segmentation proposals and the objects present in the image. We extend the composite likelihood methodology to error distributions on such one-dimensional statistical estimates, and define continuous latent variables on superpixels obtained by multiple intersections of segments, then output the optimal segments from the inferred superpixel statistics. The algorithm is capable of recombine initial mid-level proposals, as well as handle multiple interacting objects, even from the same class, all in a consistent joint inference framework by maximizing the composite likelihood of the underlying statistical model using an EM algorithm. In the PASCAL VOC segmentation challenge, the proposed approach obtains high accuracy and successfully handles images with complex object interactions.