Researchers have developed Gaze-DETR, a novel approach for detecting small, weak targets in infrared imagery, particularly for Unmanned Aerial Vehicles (UAVs). This method incorporates a bio-inspired internal priority map to guide the detection process before localization, addressing challenges like clutter and occlusion. Gaze-DETR utilizes a priority head to predict this map, enhances high-priority features with Residual Priority-Guided Feature Modulation (RPFM), and injects these priorities into decoder queries via Priority-Guided Anchor Query Injection (PAQI). The system demonstrates strong performance on new datasets like TIR-UAV120-Gaze and Anti-UAV410, showing that explicit spatial-priority learning complements traditional bounding-box supervision. AI
IMPACT Introduces a novel detection method for infrared imagery, potentially improving autonomous systems' ability to identify small targets.
RANK_REASON The cluster contains a research paper detailing a new model and dataset for computer vision. [lever_c_demoted from research: ic=1 ai=1.0]
- Anti-UAV410
- DEtection TRansformer
- Gaze-DETR
- Infrared Small Target Detection Using a Temporal Variance and Spatial Patch Contrast Filter
- Priority-Guided Anchor Query Injection
- Residual Priority-Guided Feature Modulation
- TIR-UAV120-Gaze
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
