Dr. Yingmei Zhang from the School of Software and Internet of Things Engineering Publishes High-Impact Research in IEEE TGRS
Recently,Dr. Yingmei Zhang, a young faculty member of the School of Softwareand Internet of Things Engineering, has recently published a research paper titled “HAFNet:Hierarchical Attention Fusion Network for Infrared Small Target Detection” in IEEE Transactions on Geoscience and Remote Sensing (TGRS), atop-tier and internationally recognized journal in the field of remotesensing.
The study introduces HAFNet, a novel hierarchical attention fusionnet work built upon the U-Net architecture. By integrating adual-branch semantic perception module with a hierarchical feature fusion encoder–decoder, HAFNet significantly enhances target localization capability under complex conditions and achieves substantial improvements in detection accuracy.
The research makes the following key contributions: Dual-Branch SemanticPerception Module.Designedas the backbone for feature extraction, DSPM employs standard and dilated convolutions together with spatial and channel attention mechanisms to strengthen contextual semantic interactions and effectively suppress background noise, enabling accurate small-target discrimination; Hierarchical Feature Fusion Encoder-Decoder.By extending the conventional skip connections using hierarchical attention-guided and encoded feature injection mechanisms, the proposed encoder–decoder structure enables more effective fusion ofmulti-scale and multi-level semantic features; State-of-the-ArtDetection Performance. Comprehensive experiments conducted on three benchmark datasets—NUAA-SIRST, IRSTD-1K, and NUDT-SIRST—verify that HAFNet surpasses existing infrared small target detection approaches and achieves state-of-the-art performance.
Dr.Zhang’s work has garnered considerable attention within the academic community. The proposed framework not only advances the theoretical understanding of infrared small target detection but also provides practical methodological contributions with broad application potential in remote sensing, surveillance, and aerospacefields.