ODverse33: Is the New YOLO Version Always Better? A Multi Domain benchmark from YOLO v5 to v11

TL;DR

This paper evaluates the performance of YOLO v5 to v11 across 33 datasets in 11 domains, revealing that newer versions are not always superior and highlighting the importance of domain-specific performance.

Contribution

It introduces ODverse33, a comprehensive multi-domain benchmark for YOLO models from v5 to v11, and analyzes the real-world impact of their innovations.

Findings

Newer YOLO versions do not always outperform previous ones.

Model improvements have varying effects across different domains.

The benchmark provides insights for selecting suitable YOLO versions for specific applications.

Abstract

You Look Only Once (YOLO) models have been widely used for building real-time object detectors across various domains. With the increasing frequency of new YOLO versions being released, key questions arise. Are the newer versions always better than their previous versions? What are the core innovations in each YOLO version and how do these changes translate into real-world performance gains? In this paper, we summarize the key innovations from YOLOv1 to YOLOv11, introduce a comprehensive benchmark called ODverse33, which includes 33 datasets spanning 11 diverse domains (Autonomous driving, Agricultural, Underwater, Medical, Videogame, Industrial, Aerial, Wildlife, Retail, Microscopic, and Security), and explore the practical impact of model improvements in real-world, multi-domain applications through extensive experimental results. We hope this study can provide some guidance to the extensive users of object detection models and give some references for future real-time object detector development.