… In this paper, we address three of these tasks, depth prediction, surface normal
estimation and semantic segmentation — all using a single common architecture. Our multiscale …

… Also, a multi-scale architecture approach in our task certainly improves the prediction of
semantic labels by providing a global view of image. Overall we infer that our findings aids the …

… a model that can predict surface normals and semantic labels for each pixel, … predict depth,
surface normals, and semantic segmentation using the same multi-scale network architecture …

… from multi-scale image patches to depth or surface normal … depth may be predicted for
super-pixels using regression. Now our goal is to refine the predicted depth or surface normals …

… We build a recursive search space to encode multi-scale context information for dense
prediction tasks that we term a Dense Prediction Cell (DPC). The cell is represented by a directed …

… proposes a fully convolutional deep architecture for predicting depth and surface normal from
a … [16] proposed a two-stage multi-scale CNN architecture; in the first stage, a CNN model …

… depth targets during training; however, at test time our system is purely software-based,
predicting depth … For a predicted depth map y and ground truth y∗, each with n pixels indexed by i…

… Inspired by recent works on multiscale convolutional neural networks (CNN), we propose a
deep … First, we propose a novel approach for predicting depth maps from RGB inputs which …

… This paper presents a special CNN architecture which set the state-of-the-art of three … ,
notably depth and surface normal prediction in monocular images as well as semantic labeling of …

… Our multi-scale CNN architecture for predicting the affordance labels in the indoor scenes. …
to predict the affordance labels. Thin lines represent direct input and bold lines represent a …