Computer Science · Computer Vision and Pattern Recognition
Large-scale image classification using supervised spatial encoder
Dmitriy Bespalov, Yanjun Qi, Bing Bai, Ali Shokoufandeh
See on the mapAbstract
Spatial pyramid matching (SPM) component is part of most state-of-art image classification methods. SPM encodes spatial distribution of image features, in an unsupervised fashion, by partitioning an image into regions at multiple scales and concatenating feature vectors for these regions. In this paper we propose to replace the unsupervised SPM procedure with a supervised two-stage feature selection that requires the image partitioned at a single scale. Experimental results show the proposed method performs statistically significantly better than the SPM baseline. 1.