Research map / 3D vision / Depth and stereo
Event-Intensity Stereo: Estimating Depth by the Best of Both Worlds
Mohammad Mostafavi, Kuk-Jin Yoon, Jonghyun Choi
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 intensity frames to event-only stereo | stereo network · MVSEC split 2 | mean depth error (cm): 25.18 -> 18.43 | Improved |
| add intensity frames to event-only stereo | stereo network · ECAR | mean depth error (cm): 22.3 -> 11.8 | Improved |
| add intensity frames to event-only stereo | stereo network · MVSEC split 1 | mean depth error (cm): 13.27 -> 13.74 | Hurt |
| add events to intensity-only stereo | stereo network · MVSEC split 3 | mean depth error (cm): 23.78 -> 22.36 | Improved |
| use EIS instead of the events-only baseline | stereo depth · DSEC | MAE: 0.576 -> 0.396 | Improved |
| use event-only ES variant instead of the events-only baseline | stereo depth · DSEC | MAE: 0.576 -> 0.529 | Improved |
| add feature pyramid network (ablation) | full network · ECAR subset | mean depth error (cm): 37.8 -> 8.3 | Improved |
| add deformable aggregation network (ablation) | full network · ECAR subset | mean depth error (cm): 23.4 -> 8.3 | Improved |
| add multi-scale refinement (ablation) | full network · ECAR subset | mean depth error (cm): 13.8 -> 8.3 | Improved |
| train with missing modalities | EIS with no left event stack · missing data test | mean depth error (cm): 87.5 -> 17.9 | Improved |
| train with missing modalities | EIS with full modality at test · missing data test | mean depth error (cm): 10.6 -> 16.8 | Hurt |
| use EIS instead of IS under inconsistent left-right lighting | stereo depth · inconsistent light | mean depth error (cm): 69.0 -> 32.0 | Improved |
Closest papers by meaning
- Selection and Cross Similarity for Event-Image Deep Stereo (ECCV 2022 · similarity 0.92)
- Event-Based Stereo Depth Estimation: A Survey (TPAMI 2025 · similarity 0.92)
- Stereo Depth From Events Cameras: Concentrate and Focus on the Future (CVPR 2022 · similarity 0.92)
- Zero-Shot Event-Intensity Asymmetric Stereo via Visual Prompting from Image Domain (NeurIPS 2024 · similarity 0.90)
- Active Event-based Stereo Vision (CVPR 2025 · similarity 0.89)
- Event-Driven Dynamic Scene Depth Completion (NeurIPS 2025 · similarity 0.89)
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.
