No abstract

We focus on the problem of semantic segmentation based on RGB-D data, with emphasis on analyzing cluttered indoor scenes containing many instances from many visual categories. Our approach is based on a parametric figure-ground intensity and depth-constrained proposal process that generates spatial layout hypotheses at multiple locations and scales in the image followed by a sequential inference algorithm that integrates the proposals into a complete scene estimate. Our contributions can be summarized as proposing the following:(1) a generalization of parametric max flow figure-ground proposal methodology to take advantage of intensity and depth information, in order to systematically and efficiently generate the breakpoints of an underlying spatial model in polynomial time, (2) new region description methods based on second-order pooling over multiple features constructed using both intensity and depth channels, (3) an inference procedure that can resolve conflicts in overlapping spatial partitions, and handles scenes with a large number of objects category instances, of very different scales, (4) extensive evaluation of the impact of depth, as well as the effectiveness of a large number of descriptors, both pre-designed and automatically obtained using deep learning, in a difficult RGB-D semantic segmentation problem with 92 classes. We report state of the art results in the challenging NYU Depth v2 dataset [1], extended for RMRC 2013 Indoor Segmentation Challenge, where currently the proposed model ranks first, with an average score of 24,61% and a number of 39 classes won. Moreover, we show that by combining secondorder and deep learning features, over 15% relative accuracy improvements can be additionally achieved. In a scene classification benchmark, our methodology further improves the state of the art by 24%. 1 Note that in parallel with the initial versions of our work [40,41], ideas based on our earlier RGB-based constrained parametric min cuts (CPMC) [12] and second-order pooling (O2P) [5] have also been used for RGB-D data in [39]. In any case, notice however, that [39] address the different task of 3D object detection, providing methodology to assign labels to 3D cuboids, instead of a pixel-level segmentation, as our focus in this work.