Abstract
Sparse signal models learned from data are widely used in audio, im- age, and video restoration. They have recently been generalized to discriminative image understanding tasks such as texture segmentation and feature selection. This paper extends this line of research by proposing a multiscale method to min- imize least-squares reconstruction errors and discriminative cost functions under regularization constraints. It is applied to edge detection, category-based edge selection and image classi fication tasks. Experiments on the Berkeley edge detection benchmark and the PASCAL VOC’05 and VOC’07 datasets demon- strate the computational efficiency of our algorithm and its ability to learn local image descriptions that effectively support demanding computer vision tasks.