A High-Performance Thermal Infrared Object Detection Framework with Centralized Regulation

TL;DR

This work tackles the challenge of robust object detection in Thermal Infrared (TIR) imagery by introducing CRT-YOLO, a framework that couples centralized feature regulation with efficient attention. It fuses global-range interactions via a Centralized Feature Pyramid (CFP) and long-range context with Efficient Multi-Scale Attention (EMA), while incorporating a global-to-local regulation pathway. Through extensive experiments on FLIR and LLVIP, CRT-YOLO achieves state-of-the-art performance, with ablations confirming the usefulness of EMA, EVC, and GCR components. The approach promises practical impact for nighttime surveillance, automotive safety, and industrial monitoring by delivering accurate, real-time TIR object detection with robust cross-scale representations.

Abstract

Thermal Infrared (TIR) technology involves the use of sensors to detect and measure infrared radiation emitted by objects, and it is widely utilized across a broad spectrum of applications. The advancements in object detection methods utilizing TIR images have sparked significant research interest. However, most traditional methods lack the capability to effectively extract and fuse local-global information, which is crucial for TIR-domain feature attention. In this study, we present a novel and efficient thermal infrared object detection framework, known as CRT-YOLO, that is based on centralized feature regulation, enabling the establishment of global-range interaction on TIR information. Our proposed model integrates efficient multi-scale attention (EMA) modules, which adeptly capture long-range dependencies while incurring minimal computational overhead. Additionally, it leverages the Centralized Feature Pyramid (CFP) network, which offers global regulation of TIR features. Extensive experiments conducted on two benchmark datasets demonstrate that our CRT-YOLO model significantly outperforms conventional methods for TIR image object detection. Furthermore, the ablation study provides compelling evidence of the effectiveness of our proposed modules, reinforcing the potential impact of our approach on advancing the field of thermal infrared object detection.

Figures (7)

• Figure 1: Illustration of YOLOv8 architecture.
• Figure 2: Illustration of the pipeline of our proposed CRT-YOLO.
• Figure 3: Illustration of EMA module. AP: Average Pooling; H-AP: Horizontal Average Pooling; V-AP: Vertical Average Pooling; GN: Group Normalization; $\sigma$($\cdot$): sigmoid function; sm($\cdot$): softmax function; MM: Matrix Multiplication.
• Figure 4: Illustration of Centralized Feature Pyramid (CFP) network.
• Figure 5: Illustration of lightweight MLP in the EVC module.