Hybrid Mask Generation for Infrared Small Target Detection with Single-Point Supervision
This work addresses the challenge of detecting small targets in infrared images with minimal annotation effort, offering an incremental improvement over existing weakly-supervised methods.
The paper tackled the problem of infrared small target detection with only point-level supervision by proposing a hybrid method that combines learning-free and learning-based approaches to generate high-quality pseudo masks. The method achieved a 4.3% higher average IoU than the second-best learning-free competitor and a further 3.4% increase with the hybrid approach.
Single-frame infrared small target (SIRST) detection poses a significant challenge due to the requirement to discern minute targets amidst complex infrared background clutter. In this paper, we focus on a weakly-supervised paradigm to obtain high-quality pseudo masks from the point-level annotation by integrating a novel learning-free method with the hybrid of the learning-based method. The learning-free method adheres to a sequential process, progressing from a point annotation to the bounding box that encompasses the target, and subsequently to detailed pseudo masks, while the hybrid is achieved through filtering out false alarms and retrieving missed detections in the network's prediction to provide a reliable supplement for learning-free masks. The experimental results show that our learning-free method generates pseudo masks with an average Intersection over Union (IoU) that is 4.3% higher than the second-best learning-free competitor across three datasets, while the hybrid learning-based method further enhances the quality of pseudo masks, achieving an additional average IoU increase of 3.4%.