We propose a new method for synthesizing freeviewpoint images from multiple view videos in soccer stadium. The previous method[1] can synthesize virtual viewpoint images at only i...
This paper presents research that led to the design and implementation of an extensible and scalable software framework for the dynamic 3D visualization of simulated construction ...
Abstract. This paper presents a novel approach to the problem of determining head pose estimation and face 3D orientation of several people in low resolution sequences from multipl...
Cristian Canton-Ferrer, Josep R. Casas, Montse Par...
Abstract. Object viewpoint classification aims at predicting an approximate 3D pose of objects in a scene and is receiving increasing attention. State-of-the-art approaches to view...
We describe a learning based method for recovering 3D human body pose from single images and monocular image sequences. Our approach requires neither an explicit body model nor pri...