HRYNet: A Highly Robust YOLO Network for Complex Road Traffic Object Detection

Abstract

Object detection is a crucial component of the perception system in autonomous driving. However, the road scene presents a highly intricate environment where the visibility and characteristics of traffic targets are susceptible to attenuation and loss due to various complex road scenarios such as lighting conditions, weather conditions, time of day, background elements, and traffic density. Nevertheless, the current object detection network must exhibit more learning capabilities when detecting such targets. This also exacerbates the loss of features during the feature extraction and fusion process, significantly compromising the network’s detection performance on traffic targets. This paper presents a novel methodology by which to overcome the concerns above, namely HRYNet. Firstly, a dual fusion gradual pyramid structure (DFGPN) is introduced, which employs a two-stage gradient fusion strategy to enhance the generation of more comprehensive multi-scale high-level semantic information, strengthen the interconnection between non-adjacent feature layers, and reduce the information gap that exists between them. HRYNet introduces an anti-interference feature extraction module, the residual multi-head self-attention mechanism (RMA). RMA enhances the target information by implementing a characteristic channel weighting policy, thereby reducing background interference and improving the attention capability of the network. Finally, the detection performance of HRYNet was evaluated by utilizing three datasets: the horizontally collected dataset BDD1000K, the UAV high-altitude dataset Visdrone, and a custom dataset. Experimental results demonstrate that HRYNet achieves a higher mAP_0.5 compared with YOLOv8s on the three datasets, with increases of 10.8%, 16.7%, and 5.5%, respectively. To optimize HRYNet for mobile devices, this study presents Lightweight HRYNet (LHRYNet), which effectively reduces the number of model parameters by 2 million. The results demonstrate that LHRYNet outperforms YOLOv8s in terms of mAP_0.5, with improvements of 6.7%, 10.9%, and 2.5% observed on the three datasets, respectively.

Lightweight Vehicle Detection Based on Improved YOLOv5s

“…Combined with on-board cameras and vehicle detection algorithms, the vehicle system is able to sense vehicle dynamics on the surrounding roads in real time. Moreover, vehicle detection algorithms combined with millimetre wave radar or LiDAR can improve the robustness and reliability of vehicle detection [ 56 ]. Some researchers have chosen to use the original YOLO algorithm in vehicle detection [ 4 , 5 ].…”

OD-YOLO: Robust Small Object Detection Model in Remote Sensing Image with a Novel Multi-Scale Feature Fusion

“…We also define the shape loss as ζ, with the specific calculation equation shown in Equations ( 10) and (11).…”

“…The parameter range of θ is [2,6], signifying the level of focus on shape loss, and it is set to 4 in the OD-YOLO. The expressions for W w and W h are shown in Equation (11).…”

“…Equations ( 10) and (11) measure the shape discrepancy between the predicted box and the real box. Through calculating the relative error in position of two boxes, and applying an exponential function to adjust the sensitivity of the loss, it adjusts the emphasis on shape loss.…”

“…Ding et al [ 10 ] improved the YOLOv5 model’s capacity for global feature fusion by integrating the EPMS module. Tang et al [ 11 ] improved the object detection ability of the YOLO model in complex traffic roads by introducing an RMA module. Wu et al [ 12 ] proposed the SOD-YOLO model to achieve outstanding capacity in detecting mini-goal in the high-voltage transmission lines.…”