Noise-Robust Box-Supervised Infrared Small Target Detection via Physics-Inspired Soft Label Optimization
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
The authors address the challenge of detecting small infrared targets when only rough box annotations are available, which are cheaper but less precise than pixel-level masks. They propose a method called HALO that finds a key point inside each box and creates a smooth, soft label map around it, turning noisy box labels into better guidance for learning. This method works offline, does not interfere with the detector, and improves robustness especially when the boxes are less accurate. Their experiments show HALO performs well compared to other box-supervised methods, and they analyze how noise levels affect detection performance.
Authors
Xizhe Zhang, Fan Shi, Mianzhao Wang, Jiangpeng Zheng, Xu Cheng, Shengyong Chen
Abstract
Infrared small target detection (IRSTD) commonly relies on pixel-level mask supervision. Such annotations, however, are costly and inherently uncertain because infrared targets have blurred boundaries and weak textures. We formulate box-supervised IRSTD as a problem distinct from generic box-to-mask segmentation and point-supervised IRSTD. Its central challenge is to construct stable pixel-level soft supervision from highly contaminated boxes. To this end, we propose Hotspot-Anchored Label Optimization (HALO). HALO localizes a radiometric anchor inside each box under local background-statistics constraints, then synthesizes a Physically Anchored Gaussian (PAG) soft label around the anchor. This turns noisy box supervision into continuous, pixel-level soft labels. The entire process is performed offline before training, remains decoupled from the detector backbone, and requires no online label updates. Experiments on public datasets show that HALO is competitive with representative box-supervised methods under standard tight boxes. Under looser or shifted box annotations that better approximate real scenarios, HALO is substantially more robust while remaining consistent across backbones. We further introduce a contamination-aware operating-regime analysis to characterize the effective boundary of this class of methods and reveal how intrinsic signal-to-clutter ratio relates to performance.