CA-YOLO: Cross Attention Empowered YOLO for Biomimetic Localization

2026-01-01

IEEE transactions on circuits and systems for video technology (Print)

AI Summary

This work proposes a bio-inspired stable object localization system to address the challenge of simultaneously achieving high precision and robust small-object detection in complex scenes. Built upon the YOLO architecture, the system integrates a Cross-scale Feature Fusion Attention Mechanism (CFAM) and a dedicated small-object detection head. Furthermore, it introduces a bio-inspired gimbal control strategy inspired by animal visual focus and the vestibulo-ocular reflex, enabling integrated centering, stability optimization, and intelligent reacquisition. Experimental results demonstrate that the proposed method improves average precision by 3.94% on the COCO dataset and by 4.90% on the VisDrone dataset. Additional time-sensitive target localization experiments further validate its effectiveness and practical utility.

Technology Category

• Computer Vision: Object Detection & Categorization
• Machine Learning: Bio-inspired Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

In modern complex environments, achieving accurate and efficient target localization is essential in numerous fields. However, existing systems often face limitations in both accuracy and the ability to recognize small targets. In this study, we propose a bionic stabilized localization system based on CA-YOLO, designed to enhance both target localization accuracy and small target recognition capabilities. Acting as the “brain” of the system, the target detection algorithm emulates the visual focusing mechanism of animals by integrating bionic modules into the YOLO backbone network. These modules include the introduction of a small target detection head and the development of a Characteristic Fusion Attention Mechanism (CFAM). Furthermore, drawing inspiration from the human Vestibulo-Ocular Reflex (VOR), a bionic pan-tilt tracking control strategy is developed, which incorporates central positioning, stability optimization, adaptive control coefficient adjustment, and an intelligent recapture function. The experimental results show that CA-YOLO outperforms the original model on standard datasets (COCO and VisDrone), with average accuracy metrics improved by 3.94% and 4.90%, respectively. Further time-sensitive target localization experiments validate the effectiveness and practicality of this bionic stabilized localization system.

Problem

• target localization
• small object detection
• accuracy
• complex environments
• biomimetic system

Innovation

• Cross Attention
• Biomimetic Localization
• Small Object Detection
• Characteristic Fusion Attention Mechanism
• Vestibulo-Ocular Reflex