Improved Small Object Detection Algorithm Based on YOLOv5

YOLOv5 is a popular object detection algorithm that is widely used in various industrial fields, especially in the field of autonomous driving. However, this algorithm has problems, such as false positives and false negatives when detecting small targets. The article proposes an improved method for small object detection using YOLOv5s. First, a multilevel feature fusion detection head is proposed to extract larger feature maps from the backbone of the model, improving the ability to extract features of small objects. Second, a decoupled attention mechanism is introduced at each detection head, which separates the detection of object box position, object box confidence, and class probability to reduce confusion between different feature information. Finally, the focal minimum points distance intersection over union loss function is adopted to mitigate the effects of class imbalance and poor-quality object pixels.

Authors

• Bo Xu Institute of Automation, Chinese Academy of Sciences
• Bin Gao National University of Singapore
• Yunhu Li

Keywords

• Feature extraction
• YOLO
• Semantics
• Intelligent systems
• Remote sensing
• Object detection
• Industries
• Performance evaluation
• Small Objects
• Small Object Detection
• Loss Function
• False Negative
• Convolutional Neural Network
• Aspect Ratio
• Feature Maps
• Attention Mechanism
• Intersection Over Union
• Light Signal
• Generative Adversarial Networks
• Class Probabilities
• Backbone Network
• Object Position
• Mean Average Precision
• Focus In The Field
• Position Detection
• Object Boxes
• Feature Pyramid Network
• Focal Loss
• Multi-scale Fusion
• Parameter Count
• Accurate Object Detection
• Bounding Box Regression
• Bounding Box
• Traffic Light
• Semantic Features
• Experimental Group