Two research papers propose novel approaches to infrared small target detection (IRSTD), a challenging task due to low signal-to-noise ratios and small target sizes. The first paper introduces SEF-DETR, which refines query initialization using spectral screening and frequency-routed examination to overcome limitations of standard DETR-based detectors overwhelmed by background features. The second paper, SPIRE, reformulates IRSTD as a centroid regression task, utilizing single-point supervision to guide an encoder-only framework and achieve competitive detection with reduced computational cost. AI
IMPACT These papers introduce new techniques for improving the accuracy and efficiency of infrared small target detection, potentially benefiting surveillance and autonomous systems.
RANK_REASON Two academic papers published on arXiv proposing novel methods for infrared small target detection.
- arXiv
- DETR-based detectors
- Frequency-Routed Examination
- High-Resolution Probabilistic Encoder
- Hugging Face
- Patch-wise Spectral Screening
- Point-Response Prior Supervision
- Reliability-Consistency-aware Fusion
- Rixiang Ni
- SEF-DETR
- Single-Point Supervision
- Yuteng Liu
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
