新科学想法 › 学术文库 › 学术文献 › 浏览文献

YOLOv4: Optimal Speed and Accuracy of Object Detection

handsomeland 添加于 2021-8-4 12:44 | 753 次阅读 | 0 个评论

作者
Bochkovskiy A, Wang C-Y, Liao H-YM
摘要
There are a huge number of features which are said to improve Convolutional Neural Network (CNN) accuracy. Practical testing of combinations of such features on large datasets, and theoretical justification of the result, is required. Some features operate on certain models exclusively and for certain problems exclusively, or only for small-scale datasets; while some features, such as batch-normalization and residual-connections, are applicable to the majority of models, tasks, and datasets. We assume that such universal features include Weighted-Residual-Connections (WRC), Cross-Stage-Partial-connections (CSP), Cross mini-Batch Normalization (CmBN), Self-adversarial-training (SAT) and Mish-activation. We use new features: WRC, CSP, CmBN, SAT, Mish activation, Mosaic data augmentation, CmBN, DropBlock regularization, and CIoU loss, and combine some of them to achieve state-of-the-art results: 43.5% AP (65.7% AP50) for the MS COCO dataset at a realtime speed of ~65 FPS on Tesla V100. Source code is at https://github.com/AlexeyAB/darknet
详细资料
- 关键词: cs.CV; eess.IV
- 文献种类: Manual Script
- 期卷页: 2020年
- 日期: 2020-04-23
- 发布方式: arXiv e-prints
- 备注:http://arxiv.org/abs/2004.10934v1
相关链接 URL