%0 Journal Article
%T Indoor Semantic Segmentation using depth information
%A Camille Couprie
%A Cl¨Śment Farabet
%A Laurent Najman
%A Yann LeCun
%J Computer Science
%D 2013
%I arXiv
%X This work addresses multi-class segmentation of indoor scenes with RGB-D inputs. While this area of research has gained much attention recently, most works still rely on hand-crafted features. In contrast, we apply a multiscale convolutional network to learn features directly from the images and the depth information. We obtain state-of-the-art on the NYU-v2 depth dataset with an accuracy of 64.5%. We illustrate the labeling of indoor scenes in videos sequences that could be processed in real-time using appropriate hardware such as an FPGA.
%U http://arxiv.org/abs/1301.3572v2