Hybrid Mask Generation for Infrared Small Target Detection with Single-Point Supervision

TL;DR

This paper introduces a hybrid weakly-supervised approach combining learning-free and learning-based methods to generate high-quality pseudo masks for infrared small target detection, improving accuracy over existing methods.

Contribution

It proposes a novel hybrid method that integrates sequential learning-free mask generation with filtering-based learning, enhancing pseudo mask quality in weakly-supervised infrared small target detection.

Findings

Learning-free method achieves 4.3% higher IoU than competitors.

Hybrid approach further improves IoU by 3.4%.

Method demonstrates superior performance across three datasets.

Abstract

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%.