Research map / 3D vision / Depth and stereo
Enhanced Event-based Dense Stereo via Cross-Sensor Knowledge Distillation
Haihao Zhang, Yunjian Zhang, Jianing Li, Lin Zhu, Meng Lv, Yao Zhu, Yanwei Liu, Xiangyang Ji
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 feature distillation (shallow and deep) (ablation) | IED2S · MVSEC split1 | mean disparity error (pix): 0.54 -> 0.38 | Improved |
| add feature distillation (ablation) | IED2S · MVSEC split1/split3 | one-pixel accuracy: 91.0 -> 94.4 | Improved |
| add logit-level distillation alone (ablation) | IED2S · MVSEC split1 | mean disparity error (pix): 0.54 -> 0.49 | Improved |
| add left-right consistency alone (ablation) | IED2S · MVSEC split3 | one-pixel accuracy: 89.4 -> 90.3 | Improved |
| add LRC on top of F-distill (ablation) | IED2S · MVSEC split1 | mean disparity error (pix): 0.38 -> 0.37 | Improved |
| add logit distillation on top of F-distill and LRC (ablation) | IED2S · MVSEC split1 | mean disparity error (pix): 0.37 -> 0.32 | Improved |
| add all components (F-distill, L-distill, LRC) (ablation) | IED2S · MVSEC split3 | one-pixel accuracy: 89.4 -> 93.4 | Improved |
| use proposed shallow distillation instead of plain MSE feature alignment (ablation) | IED2S · MVSEC split1 | mean disparity error (pix): 0.45 -> 0.39 | Improved |
| add deep distillation to shallow distillation (ablation) | IED2S · MVSEC split1 | mean disparity error (pix): 0.39 -> 0.32 | Improved |
| add distillation (full model vs none) (ablation) | IED2S · MVSEC -> DSEC | MAE: 0.782 -> 0.563 | Improved |
| add distillation (full model vs none) (ablation) | IED2S · DSEC -> MVSEC | one-pixel accuracy: 79.8 -> 91.4 | Improved |
| use intensity-to-event distillation vs best event-only method TES | IED2S · MVSEC split1 | mean depth error (cm): 13.0 -> 11.2 | Improved |
Closest papers by meaning
- Deep Event Stereo Leveraged by Event-to-Image Translation (AAAI 2021 · similarity 0.91)
- Learning an Event Sequence Embedding for Dense Event-Based Deep Stereo (ICCV 2019 · similarity 0.91)
- Temporal Stereo Matching From Event Cameras via Joint Learning With Stereoscopic Flow (TPAMI 2025 · similarity 0.90)
- Temporal Event Stereo via Joint Learning with Stereoscopic Flow (ECCV 2024 · similarity 0.90)
- Active Event-based Stereo Vision (CVPR 2025 · similarity 0.88)
- Zero-Shot Event-Intensity Asymmetric Stereo via Visual Prompting from Image Domain (NeurIPS 2024 · similarity 0.88)
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.
