We study the notion of consistency between a 3D shape and a 2D observation and propose a differentiable formulation which allows computing gradients of the 3D shape given an observation from an arbitrary view. We do so by reformulating view consistency using a differentiable ray consistency (DRC) term. We show that this formulation can be incorporated in a learning framework to leverage different types of multi-view observations e.g. foreground masks, depth, color images, semantics etc. as supervision for learning single-view 3D prediction. We present empirical analysis of our technique in a controlled setting. We also show that this approach allows us to improve over existing techniques for single-view reconstruction of objects from the PASCAL VOC dataset.

No abstract

No abstract

No abstract

No abstract

Any sensory system can be viewed as a passive or dumb element which provides raw data. It can also be viewed as an intelligent element which returns "analyzed" information. Finally, it can be viewed as a commanding element which sends commands to the physical system. Each of these views is used in different situations and for different tasks. Commanding sensors are an extension to the logical sensor approach in which a mapping from events to actions is added to the sensor model.In a previous paper, we proposed a sensor-based distributed control scheme for mobile robots along with several simulation results [ 11. In this paper, the application of this scheme to control a real mobile robot is presented and the results of several experiments are discussed. A server-client model is used to implement this scheme where the server is a process that carries out the commands to be executed, and each client is a process with a certain task. The logical sensor approach is used to model the sensory system which provides different levels of data representation with tolerance measures and analysis.

No abstract