Research map / Image generation / Autoregressive and latent generation
AV2T-Gen: Aerial Visible to Thermal Generation with Environment and Vehicle State Guidance
Kun Yang, Yuxiang Liu, Yihan Wang, Shen Yan, Maojun Zhang, Yu Liu, Xue Wang, Qing Wang
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 environment prompt (ablation) | AV2T-Gen · Aerial RGB-T (All scenes) | PSNR: 22.29 -> 24.08 | Improved |
| add vehicle state labels (ablation) | AV2T-Gen · Aerial RGB-T (All scenes) | PSNR: 22.93 -> 24.08 | Improved |
| add edge loss (ablation) | AV2T-Gen · Aerial RGB-T (All scenes) | PSNR: 22.90 -> 24.08 | Improved |
| add perceptual loss (ablation) | AV2T-Gen · Aerial RGB-T (All scenes) | PSNR: 23.14 -> 24.08 | Improved |
| add temperature to environment prompt (ablation) | AV2T-Gen · Aerial RGB-T (All scenes) | PSNR: 23.41 -> 24.08 | Improved |
| add humidity to environment prompt (ablation) | AV2T-Gen · Aerial RGB-T (All scenes) | PSNR: 23.68 -> 24.08 | Improved |
| add wind speed to environment prompt (ablation) | AV2T-Gen · Aerial RGB-T (All scenes) | PSNR: 23.85 -> 24.08 | Improved |
| add air pressure to environment prompt (ablation) | AV2T-Gen · Aerial RGB-T (All scenes) | PSNR: 23.97 -> 24.08 | Improved |
| train on three scenes and test on unseen fourth scene (leave-one-out) | AV2T-Gen · Aerial RGB-T Scene 1 | PSNR: 25.27 -> 23.75 | Hurt |
| zero-shot transfer to LLVIP versus ThermalGen | AV2T-Gen · LLVIP | PSNR: 12.58 -> 13.82 | Improved |
| use AV2T-Gen over ThermalGen for synthetic thermal data | AV2T-Gen · Vehicle detection on generated thermal images | mAP (%): 11.7 -> 15.6 | Improved |
Closest papers by meaning
- Visible-Thermal UAV Tracking: A Large-Scale Benchmark and New Baseline (CVPR 2022 · similarity 0.86)
- Diagnosing Aerial-View Object Detectors with Foundational Image Generative Models (ECCV 2026 · similarity 0.85)
- TrafficNight : An Aerial Multimodal Benchmark For Nighttime Vehicle Surveillance (ECCV 2024 · similarity 0.85)
- ThermalGen: Style-Disentangled Flow-Based Generative Models for RGB-to-Thermal Image Translation (NeurIPS 2025 · similarity 0.85)
- VVSim: A Large-Scale Aerial-Ground Dataset and Benchmark for Cooperative Perception (ECCV 2026 · similarity 0.85)
- Caltech Aerial RGB-Thermal Dataset in the Wild (ECCV 2024 · 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.
