Boosting-like Deep Learning For Pedestrian Detection
This work addresses overfitting in pedestrian detection, an incremental improvement for computer vision applications.
The paper tackles overfitting in deep learning for pedestrian detection by incorporating a boosting-like technique to weigh training samples, achieving a 15.85% and 3.81% reduction in average miss rate compared to ACF and JointDeep on the Caltech benchmark dataset.
This paper proposes boosting-like deep learning (BDL) framework for pedestrian detection. Due to overtraining on the limited training samples, overfitting is a major problem of deep learning. We incorporate a boosting-like technique into deep learning to weigh the training samples, and thus prevent overtraining in the iterative process. We theoretically give the details of derivation of our algorithm, and report the experimental results on open data sets showing that BDL achieves a better stable performance than the state-of-the-arts. Our approach achieves 15.85% and 3.81% reduction in the average miss rate compared with ACF and JointDeep on the largest Caltech benchmark dataset, respectively.