Abstract:Detection of tiny object in complex environments is a matter of urgency, not only because of the high real-world demand, but also the high deployment and real-time requirements. Although many current single-stage algorithms have good detection performance under low computing power requirements, there are still significant challenges such as distinguishing the background from object features and extracting small-scale target features in complex natural environments. To address this, we first created real datasets based on natural environments and improved dataset diversity using a combination of copy–paste enhancement and multiple image enhancement techniques. As for the choice of network, we chose YOLOV5s due to its nature of fewer parameters and easier deployment in the same class of models. Most improvement strategies to boost detection performance claim to improve the performance of privilege extraction and recognition. However, we prefer to consider the combination of realistic deployment feasibility and detection performance. Therefore, based on the hottest improvement methods of YOLOV5s, we try to make adaptive improvements in three aspects, namely attention mechanism, head network, and backbone network. The experimental results proved that the decoupled head and Slimneck based improvements achieved, respectively, 0.872 and 0.849, 0.538 and 0.479, 87.5% and 89.8% on the mAP0.5, mAP0.5:0.95, and Precision metrics, surpassing the results of the baseline model on these three metrics: 0.705, 0.405 and 83.6%. This result suggests that the adaptively improved model can better meet routine testing needs without significantly increasing the number of parameters. These models perform well on our custom dataset and are also effective on images that are difficult to detect by naked eye. Meanwhile, we find that YOLOV8s, which also has the decoupled head improvement, has the results of 0.743, 0.461, and 87.17% on these three metrics. It proves that under our dataset, it is possible to achieve more advanced results with lower number of model parameters just by adding decoupled head. And according to the results, we also discuss and analyze some improvements that are not adapted to our dataset, which also provides ideas for researchers in similar scenarios: in the booming development of object detection, choosing the suitable model and adapting to combine with other technologies would help to provide solutions to real-world problems.

YOLO-G: Improved YOLO for cross-domain object detection

Tri-Flow-YOLO: Counter helps to improve cross-domain object detection

Multiscale Domain Adaptive YOLO for Cross-Domain Object Detection

YOLO-SDH: improved YOLOv5 using scaled decoupled head for object detection

YOLOv10: Real-Time End-to-End Object Detection

YOLO-Extract: Improved YOLOv5 for Aircraft Object Detection in Remote Sensing Images

MS-YOLO: integration-based multi-subnets neural network for object detection in aerial images

YOLO Adaptive Developments in Complex Natural Environments for Tiny Object Detection

A YOLO-NL object detector for real-time detection

An Improved YOLOv8 Detector for Multi-Scale Target Detection in Remote Sensing Images

SSDA-YOLO: Semi-supervised domain adaptive YOLO for cross-domain object detection

YOLO-TLA: An Efficient and Lightweight Small Object Detection Model based on YOLOv5

YOLO_SRv2: An evolved version of YOLO_SR

YOLO-HR: Improved YOLOv5 for Object Detection in High-Resolution Optical Remote Sensing Images

Improved small-object detection using YOLOv8: A comparative study

A YOLOX Object Detection Algorithm Based on Bidirectional Cross-scale Path Aggregation

WGS-YOLO: A real-time object detector based on YOLO framework for autonomous driving

DS-YOLOv8-Based Object Detection Method for Remote Sensing Images

Multi-scene small object detection with modified YOLOv4

YOLO-A2G: An air-to-ground high-precision object detection algorithm based on YOLOv5

YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications