Research map / Vision-language and multimodal / Cross-modal learning
Infrared-Privileged UAV Detection via Cross-Modal Vector-Quantization
Zhibo Lou, Ruijie Zhang, Zeyu Luo, Qianxi Cao, Feng Qian, Junjie Chen, Yuming Fang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Hallu-branch network layers only (ablation) | YOLOv8m · DroneVehicle (YOLOv8m ablation) | mAP@0.5: 60.2 -> 60.8 | No clear effect |
| hallucinate IR codebook indices with CE loss (ablation) | YOLOv8m+Hallu-branch · DroneVehicle (YOLOv8m ablation) | mAP@0.5: 60.8 -> 65.7 | Improved |
| hierarchical residual quantization of multi-scale indices (ablation) | YOLOv8m+index hallucination · DroneVehicle (YOLOv8m ablation) | mAP@0.5: 65.7 -> 67.6 | Improved |
| apply CMVQ to RemDet | RemDet · DroneVehicle (RGB test input) | mAP@0.5: 65.9 -> 71.3 | Improved |
| apply CMVQ to YOLOv8m | YOLOv8m · DroneVehicle (RGB test input) | mAP@0.5: 63.9 -> 67.6 | Improved |
| apply CMVQ to YOLOv8m | YOLOv8m · VisDrone val (pretrain DroneVehicle, fine-tune) | AVG mAP: 52.0 -> 62.7 | Improved |
| hallucinate codebook indices instead of features | YOLOv8m · DroneVehicle (RGB test input) | mAP@0.5: 65.3 -> 67.6 | Improved |
| IR-privileged training vs RGB-IR test input | Ours+RemDet · DroneVehicle (RGB test input) | mAP@0.5: 71.3 vs 81.4 | Mixed |
Closest papers by meaning
- "Translation, Scale and Rotation: Cross-Modal Alignment Meets RGB-Infrared Vehicle Detection" (ECCV 2022 · similarity 0.89)
- Tri-Modal Fusion Transformers for UAV-based Object Detection (CVPR 2026 · similarity 0.88)
- Fusion Meets Diverse Conditions: A High-diversity Benchmark and Baseline for UAV-based Multimodal Object Detection with Condition Cues (ICCV 2025 · similarity 0.87)
- Weakly Misalignment-free Adaptive Feature Alignment for UAVs-based Multimodal Object Detection (CVPR 2024 · similarity 0.86)
- Detection-Friendly Nonuniformity Correction: A Union Framework for Infrared UAV Target Detection (CVPR 2025 · similarity 0.85)
- UCM-VeID V2: A Richer Dataset and A Pre-training Method for UAV Cross-Modality Vehicle Re-Identification (CVPR 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
