k-fold Subsampling based Sequential Backward Feature Elimination

2016-02-24

International Conference on Pattern Recognition Applications and Methods

AI Summary

To address the redundancy and insufficient discriminability of shape features in human detection, this paper proposes a novel feature selection method integrating filter- and wrapper-based strategies. We innovatively design a k-fold subsampling-driven Sequential Backward Elimination (SBE) framework that simultaneously optimizes feature robustness and discriminability while preserving essential shape representation capability, thereby significantly reducing feature dimensionality. Coupled with a linear SVM classifier, our method is evaluated on the INRIA and ETH pedestrian datasets. Results show that it achieves over 50% faster detection speed and a 2% improvement in mean Average Precision (mAP) compared to the current state-of-the-art approaches; relative to the Deformable Parts Model (DPM), it yields approximately a 9% mAP gain. The proposed framework establishes an interpretable, reusable paradigm for lightweight and efficient human detection through principled feature selection.

Technology Category

• Machine Learning: Dimensionality Reduction/Feature Selection
• Humans and AI: Human-in-the-loop Machine Learning
• Computer Vision: Object Detection & Categorization

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

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.

Problem

• Develops a hybrid feature selection algorithm for human detection.
• Combines k-fold subsampling and sequential backward elimination methods.
• Improves SVM classifier speed and accuracy in pedestrian detection.

Innovation

• Hybrid feature selection combining filter and wrapper methods
• Uses k-fold subsampling and sequential backward elimination
• Improves SVM detection speed by over 50%