Research map / 3D vision / Depth and stereo
LiDAR-Event Stereo Fusion with Hallucinations
Luca Bartolomei, Matteo Poggi, Andrea Conti, Stefano Mattoccia
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 |
|---|---|---|---|
| apply Guided (cost volume modulation) to pre-trained event stereo backbones | event stereo backbones · DSEC test | 1PE: 16.21 -> 16.07 (Histogram) | No clear effect |
| apply VSH or BTH hallucination to pre-trained backbones | event stereo backbones · DSEC test | 1PE: 16.21 -> 13.32 (Histogram, BTH) | Improved |
| train backbones with VSH or BTH hallucination | event stereo backbones · DSEC test (retrained) | 1PE: 16.21 -> 9.90 (Histogram, VSH) | Improved |
| apply VSH or BTH to DSEC-trained backbone evaluated on M3ED | event stereo backbones · M3ED (pre-trained) | 1PE: 37.70 -> 20.19 (Histogram, VSH) | Improved |
| use repeated injection instead of single injection in BTH (ablation) | BTH · M3ED time-misaligned LiDAR | 1PE | Improved |
| increase LiDAR time misalignment up to 100 ms (ablation) | VSH and BTH · M3ED time-misaligned LiDAR | 1PE | Hurt |
| use 3x3 patches instead of single pixel in VSH (ablation) | VSH · DSEC search split | 1PE | Improved |
| use alpha blending 0.5 in VSH (ablation) | VSH · DSEC search split | 1PE | Improved |
| inject two fictitious events per depth point instead of one in BTH (ablation) | BTH · DSEC search split | 1PE | Improved |
| use uniform polarities and uniform patches in BTH (ablation) | BTH · DSEC search split | 1PE | Improved |
| compare Concat and Guided baselines trained on DSEC, tested on M3ED | Concat / Guided fusion · M3ED (retrained) | 1PE | Mixed |
Closest papers by meaning
- Stereo Depth From Events Cameras: Concentrate and Focus on the Future (CVPR 2022 · similarity 0.89)
- Active Event-based Stereo Vision (CVPR 2025 · similarity 0.88)
- Toward Ultrafast Depth Sensing via Active Event-Based Stereo Vision (TPAMI 2026 · similarity 0.88)
- Zero-Shot Event-Intensity Asymmetric Stereo via Visual Prompting from Image Domain (NeurIPS 2024 · similarity 0.85)
- Event-Driven Dynamic Scene Depth Completion (NeurIPS 2025 · similarity 0.85)
- DERD-Net: Learning Depth from Event-based Ray Densities (NeurIPS 2025 · 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.
