Year: 2024
Type: conference-paper
Abstract: In recent years, with the rapid development of deep learning technology, small object detection methods based on deep learning have been widely studied, and deep learning algorithms can achieve good results in small object detection. For small targets in images, existing detection methods still suffer from weak feature expression ability and low detection performance. In order to improve the detection ability of small targets, this paper improves the YOLOv8 algorithm. Firstly, the BiFPN network structure is introduced in the Neck module of YOLOv8 to alleviate the problem of insufficient feature information of small targets. Then, a Coordinate Attention mechanism is added to the backbone network to focus on key features. Then, a P2 detection head is added to the Head module of the YOLOv8 network, effectively improving the serious problem of small target missed detection. Finally, in the improved loss function module, this paper adopts WiseIoU Loss to solve the problem of slow convergence speed and low efficiency of the model, further improving the overall performance of the model. After a series of experiments, the results show that the improved algorithm in this paper can effectively improve the network's ability to detect small targets.
Authors: Haojin Zhang, Wei Shi
Institution: Ningxia University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Awards: 关于双定型前向神经网络逼近能力的研究, G7869105468
OpenAlex ID: W4407601598