N-RPN: Hard Example Learning for Region Proposal Networks
This addresses a specific bottleneck in object detection for computer vision researchers, but it is incremental as it builds upon existing RPN methods.
The paper tackles the problem of region proposal networks struggling with hard negative examples, which leads to false positives and poor performance; by introducing a Negative Region Proposal Network (nRPN) that learns from false positives to provide hard negatives, it achieves performance improvements on the PASCAL VOC 2007 dataset.
The region proposal task is to generate a set of candidate regions that contain an object. In this task, it is most important to propose as many candidates of ground-truth as possible in a fixed number of proposals. In a typical image, however, there are too few hard negative examples compared to the vast number of easy negatives, so region proposal networks struggle to train on hard negatives. Because of this problem, networks tend to propose hard negatives as candidates, while failing to propose ground-truth candidates, which leads to poor performance. In this paper, we propose a Negative Region Proposal Network(nRPN) to improve Region Proposal Network(RPN). The nRPN learns from the RPN's false positives and provide hard negative examples to the RPN. Our proposed nRPN leads to a reduction in false positives and better RPN performance. An RPN trained with an nRPN achieves performance improvements on the PASCAL VOC 2007 dataset.