YOLOv4: Optimal Speed and Accuracy of Object Detection

TL;DR

This paper introduces YOLOv4, an object detection model that combines various new and existing features to achieve state-of-the-art accuracy and speed on the MS COCO dataset, suitable for real-time applications.

Contribution

The paper presents YOLOv4, a novel object detection system that integrates multiple advanced features for improved accuracy and efficiency, optimized for real-time performance.

Findings

Achieves 43.5% AP on MS COCO dataset.

Operates at approximately 65 FPS on Tesla V100.

Outperforms previous models in speed and accuracy.

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