k-fold Subsampling based Sequential Backward Feature Elimination
This work addresses feature selection for human detection in computer vision, offering incremental improvements in speed and accuracy for pedestrian datasets.
The paper tackled human detection by proposing a hybrid feature selection algorithm combining k-fold subsampling and sequential backward elimination with an SVM classifier, resulting in over 50% faster detection speed and up to 2% better accuracy compared to other methods, with a 9% accuracy improvement over deformable part models.
We present a new wrapper feature selection algorithm for human detection. This algorithm is a hybrid feature selection approach combining the benefits of filter and wrapper methods. It allows the selection of an optimal feature vector that well represents the shapes of the subjects in the images. In detail, the proposed feature selection algorithm adopts the k-fold subsampling and sequential backward elimination approach, while the standard linear support vector machine (SVM) is used as the classifier for human detection. We apply the proposed algorithm to the publicly accessible INRIA and ETH pedestrian full image datasets with the PASCAL VOC evaluation criteria. Compared to other state of the arts algorithms, our feature selection based approach can improve the detection speed of the SVM classifier by over 50% with up to 2% better detection accuracy. Our algorithm also outperforms the equivalent systems introduced in the deformable part model approach with around 9% improvement in the detection accuracy.