PP-YOLOv2: A Practical Object Detector

TL;DR

PP-YOLOv2 is an improved, practical object detector that enhances performance significantly while maintaining high speed, making it suitable for real-world applications.

Contribution

The paper introduces PP-YOLOv2, a new version of PP-YOLO that combines multiple refinements to boost accuracy without sacrificing inference speed.

Findings

Improved mAP from 45.9% to 49.5% on COCO2017 test-dev.

Achieves 68.9 FPS at 640x640 input, up to 106.5 FPS with optimizations.

Surpasses similar models like YOLOv4-CSP and YOLOv5l in performance.

Abstract

Being effective and efficient is essential to an object detector for practical use. To meet these two concerns, we comprehensively evaluate a collection of existing refinements to improve the performance of PP-YOLO while almost keep the infer time unchanged. This paper will analyze a collection of refinements and empirically evaluate their impact on the final model performance through incremental ablation study. Things we tried that didn't work will also be discussed. By combining multiple effective refinements, we boost PP-YOLO's performance from 45.9% mAP to 49.5% mAP on COCO2017 test-dev. Since a significant margin of performance has been made, we present PP-YOLOv2. In terms of speed, PP-YOLOv2 runs in 68.9FPS at 640x640 input size. Paddle inference engine with TensorRT, FP16-precision, and batch size = 1 further improves PP-YOLOv2's infer speed, which achieves 106.5 FPS. Such a performance surpasses existing object detectors with roughly the same amount of parameters (i.e., YOLOv4-CSP, YOLOv5l). Besides, PP-YOLOv2 with ResNet101 achieves 50.3% mAP on COCO2017 test-dev. Source code is at https://github.com/PaddlePaddle/PaddleDetection.