Rich feature hierarchies for accurate object detection and semantic segmentation
"Rich feature hierarchies for accura..." refers background in this paper
...The only class-specific computations are dot products between features and SVM weights and non-maximum suppression....
"Rich feature hierarchies for accura..." refers methods in this paper
...This paper is the first to show that a CNN can lead to dramatically higher object detection performance on PASCAL VOC as compared to systems based on simpler HOG-like features....
...Compared with DPM (see ), significantly more of our errors result from poor localization, rather than confusion with background or other object classes, indicating that the CNN features are much more discriminative than HOG....
...After fine-tuning, our system achieves a mAP of 54% on VOC 2010 compared to 33% for the highly-tuned, HOG-based deformable part model (DPM) [17, 20]....
...SIFT and HOG are blockwise orientation histograms, a representation we could associate roughly with complex cells in V1, the first cortical area in the primate visual pathway....
...This finding suggests potential utility in computing a dense feature map, in the sense of HOG, of an arbitrary-sized image by using only the convolutional layers of the CNN....