OHEM算法详解——online hard example mining
OHEM: online hard example mining
论文链接：https://arxiv.org/abs/1604.03540
主要特点：
1.不需要设置正负样本比例来解决数据类别不平衡问题
2.数据集越大，性能越明显
文中所说：
1.OHEM is a simple and intuitive algorithm that eliminates several heuristics and hyperparameters in common use. But more importantly, it yields consistent and significant boosts in detection performance on benchmarks like PASCAL VOC 2007 and 2012.
2.Its effectiveness increases as datasets become larger and more difficult, as demonstrated by the results on the MS COCO dataset.
算法详解：
本文来自互联网用户投稿，文章观点仅代表作者本人，不代表本站立场，不承担相关法律责任。如若转载，请注明出处。 如若内容造成侵权/违法违规/事实不符，请点击【内容举报】进行投诉反馈！