Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.02837 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917186741731328 |
|---|---|
| author | Liu, Yuteng Meng, Duanni Yuan, Maoxun Wei, Xingxing |
| author_facet | Liu, Yuteng Meng, Duanni Yuan, Maoxun Wei, Xingxing |
| contents | Infrared small target detection (IRSTD) faces significant challenges due to the low signal-to-noise ratio (SNR), small target size, and complex cluttered backgrounds. Although recent DETR-based detectors benefit from global context modeling, they exhibit notable performance degradation on IRSTD. We revisit this phenomenon and reveal that the target-relevant embeddings of IRST are inevitably overwhelmed by dominant background features due to the self-attention mechanism, leading to unreliable query initialization and inaccurate target localization. To address this issue, we propose SEF-DETR, a novel framework that refines query initialization for IRSTD. Specifically, SEF-DETR consists of three components: Frequency-guided Patch Screening (FPS), Dynamic Embedding Enhancement (DEE), and Reliability-Consistency-aware Fusion (RCF). The FPS module leverages the Fourier spectrum of local patches to construct a target-relevant density map, suppressing background-dominated features. DEE strengthens multi-scale representations in a target-aware manner, while RCF further refines object queries by enforcing spatial-frequency consistency and reliability. Extensive experiments on three public IRSTD datasets demonstrate that SEF-DETR achieves superior detection performance compared to state-of-the-art methods, delivering a robust and efficient solution for infrared small target detection task. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_02837 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Breaking Self-Attention Failure: Rethinking Query Initialization for Infrared Small Target Detection Liu, Yuteng Meng, Duanni Yuan, Maoxun Wei, Xingxing Computer Vision and Pattern Recognition Artificial Intelligence Infrared small target detection (IRSTD) faces significant challenges due to the low signal-to-noise ratio (SNR), small target size, and complex cluttered backgrounds. Although recent DETR-based detectors benefit from global context modeling, they exhibit notable performance degradation on IRSTD. We revisit this phenomenon and reveal that the target-relevant embeddings of IRST are inevitably overwhelmed by dominant background features due to the self-attention mechanism, leading to unreliable query initialization and inaccurate target localization. To address this issue, we propose SEF-DETR, a novel framework that refines query initialization for IRSTD. Specifically, SEF-DETR consists of three components: Frequency-guided Patch Screening (FPS), Dynamic Embedding Enhancement (DEE), and Reliability-Consistency-aware Fusion (RCF). The FPS module leverages the Fourier spectrum of local patches to construct a target-relevant density map, suppressing background-dominated features. DEE strengthens multi-scale representations in a target-aware manner, while RCF further refines object queries by enforcing spatial-frequency consistency and reliability. Extensive experiments on three public IRSTD datasets demonstrate that SEF-DETR achieves superior detection performance compared to state-of-the-art methods, delivering a robust and efficient solution for infrared small target detection task. |
| title | Breaking Self-Attention Failure: Rethinking Query Initialization for Infrared Small Target Detection |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2601.02837 |