vix.ing · top · new · best · stats

YOLOv4: Optimal Speed and Accuracy of Object Detection

2020/04/23 by Alexey Bochkovskiy, Chien-Yao Wang, Bochkovskiy, Alexey +3 · 2 voices · 10,479 citations
Computer Science · Engineering · #Advanced Neural Network Applications #Artificial intelligence #CCD and CMOS Imaging Sensors #Computer science #Computer vision #Industrial Vision Systems and Defect Detection #Object (grammar) #cs.CV #eess.IV

paper · pdf · doi:10.48550/arxiv.2004.10934

published in arXiv (Cornell University) (Cornell University)

arxiv created 2020/04/23 · openalex publication_date 2020/04/23 · arxiv updated 2020/04/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/02

Abstract

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

Citations

Cited by

Discussions

Related