This paper proposes a novel method for detection and segmentation of foreground objects from a video which contains both stationary and moving background objects and undergoes bot...
— Object recognition is a challenging problem for artificial systems. This is especially true for objects that are placed in cluttered and uncontrolled environments. To challenge...
Most standard learning algorithms, such as Logistic Regression (LR) and the Support Vector Machine (SVM), are designed to deal with i.i.d. (independent and identically distributed...
Chi-Hoon Lee, Matthew R. G. Brown, Russell Greiner...
While discriminative training (e.g., CRF, structural SVM) holds much promise for machine translation, image segmentation, and clustering, the complex inference these applications ...
This paper presents a statistical model for textures that uses a non-negative decomposition on a set of local atoms learned from an exemplar. This model is described by the varianc...