Local coordinate coding has recently been introduced to learning visual feature dictionary and achieved top level performance for object recognition. However, the computational co...
Recently, sparse coding has been receiving much attention in object and scene recognition tasks because of its superiority in learning an effective codebook over k-means clusterin...
This work proposes to learn visual encodings of attention patterns that enables sequential attention for object detection in real world environments. The system embeds a saccadic d...
Visual codebook based quantization of robust appearance descriptors extracted from local image patches is an effective means of capturing image statistics for texture analysis and...