By combining histogram of oriented gradient and histograms of shearlet coefficients, which analyzes images at multiple scales and orientations based on shearlet transforms, as the feature set, we proposed a novel hama...By combining histogram of oriented gradient and histograms of shearlet coefficients, which analyzes images at multiple scales and orientations based on shearlet transforms, as the feature set, we proposed a novel haman detection feature. We employ partial least squares analysis, an efficient dimensionality reduction technique, to project the feature onto a much lower dimensional subspace. We test it in INRIA person dataset by using a linear SVM, and it yields an error rate of 1.38% with a false negatives (FN) rate of 0.40% and a false positive (FP) rate of 0.98%, while the error rate of HOG is 7.11%, with a FN rate of 4.09% and a FP rate of 3.02%.展开更多
基金Supported by the National Natural Science Foundations of China (No. 90820306)the natural science foundation of Jiangsu Province Youth Fund(No. BK2012399)
文摘By combining histogram of oriented gradient and histograms of shearlet coefficients, which analyzes images at multiple scales and orientations based on shearlet transforms, as the feature set, we proposed a novel haman detection feature. We employ partial least squares analysis, an efficient dimensionality reduction technique, to project the feature onto a much lower dimensional subspace. We test it in INRIA person dataset by using a linear SVM, and it yields an error rate of 1.38% with a false negatives (FN) rate of 0.40% and a false positive (FP) rate of 0.98%, while the error rate of HOG is 7.11%, with a FN rate of 4.09% and a FP rate of 3.02%.