Research map / 3D vision / Depth and stereo
Depth-guided Multi-view Exposure Bracketing for HDR Robot Vision
Jinnyeong Kim, Juhyung Choi, Woohyeok Kim, Sunghyun Cho, Seung-Hwan Baek
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 |
|---|---|---|---|
| use monocular depth instead of optical flow for alignment (ablation) | DMEB · Real modest DR, single-shot multi-camera | PSNR-mu: 35.84 -> 37.09 | Improved |
| use single-view LiDAR-anchored depth instead of monocular depth (ablation) | DMEB · Real modest DR, single-shot multi-camera | PSNR-mu: 37.09 -> 37.90 | Improved |
| use multi-view LiDAR-guided fusion instead of single-view LiDAR depth (ablation) | DMEB · Real modest DR, single-shot multi-camera | PSNR-mu: 37.90 -> 39.40 | Improved |
| remove depth and align with optical flow (ablation) | DMEB · Real modest DR | PSNR-mu: 39.40 -> 35.84 | Hurt |
| switch from multi-shot single-camera to single-shot multi-camera input | HDR Transformer · Real modest DR | PSNR-mu: 31.26 -> 33.61 | Improved |
| switch from multi-shot single-camera to single-shot multi-camera input | AFUNet · Real modest DR | PSNR-mu: 29.64 -> 32.25 | Improved |
| switch from multi-shot single-camera to single-shot multi-camera input | DMEB · Real modest DR | PSNR-mu: 30.72 -> 39.40 | Improved |
| increase camera count from three to eight (ablation) | DMEB · Real and synthetic | dynamic range | Improved |
| use DMEB instead of HDR Transformer on ultra DR scenes | DMEB · Real ultra DR, single-shot multi-camera | HDR-VDP: PSNR-mu 38.74 -> 39.17, HDR-VDP 9.40 -> 9.16 | Mixed |
| use DMEB instead of AFUNet on synthetic benchmark | DMEB · Synthetic benchmark (CARLA) | PSNR-mu: 33.52 -> 34.72 | Improved |
| use DMEB on zero-shot external two-view data | DMEB · Choi 2-view | PSNR-mu: 26.33 and 25.39 -> 33.00 | Improved |
| use DMEB output for YOLOv8 detection | DMEB · YOLOv8 detection, 90 annotated frames | Recall/F1: 0.78/0.63 and 0.81/0.67 -> 0.83/0.73 | Improved |
Closest papers by meaning
- Dual Exposure Stereo for Extended Dynamic Range 3D Imaging (CVPR 2025 · similarity 0.87)
- InstantHDR: Single-forward Gaussian Splatting for High Dynamic Range 3D Reconstruction (ECCV 2026 · similarity 0.87)
- High Dynamic Range Novel View Synthesis with Single Exposure (ICML 2025 · similarity 0.86)
- Single Shot HDR Recovery via a Video Diffusion Prior (NeurIPS 2026 · similarity 0.85)
- HDR-GS: Efficient High Dynamic Range Novel View Synthesis at 1000x Speed via Gaussian Splatting (NeurIPS 2024 · similarity 0.85)
- Towards Co-Evaluation of Cameras HDR and Algorithms for Industrial-Grade 6DoF Pose Estimation (CVPR 2024 · similarity 0.85)
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.
