Research map / 3D vision / Depth and stereo
Towards Robust Event-Based Depth Estimation: Bridging Synthetic and Real Domains with Motion Adaptation
Yuzhe Ji, Haotian Wang, Yijie Chen, Xiang Cheng, Liuqing Yang, Xinhu Zheng
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 Spatial-Temporal Association module (ablation) | MA-Mamba · MVSEC zero-shot | Abs. Rel: 0.50 -> 0.45 | Improved |
| add Adaptive Memory Balancing module alone (ablation) | MA-Mamba · MVSEC zero-shot | Abs. Rel: 0.50 -> 0.47 | Improved |
| add event-oriented augmentations to STA+AMB (ablation) | MA-Mamba · MVSEC zero-shot | Abs. Rel: 0.40 -> 0.31 | Improved |
| add random noise injection (ablation) | MA-Mamba augmentations · MVSEC zero-shot | Abs. Rel: 0.40 -> 0.38 | No clear effect |
| add priority-based masked sequence modeling (ablation) | MA-Mamba augmentations · MVSEC zero-shot | Abs. Rel: 0.38 -> 0.35 | Improved |
| add random motion pattern generation (ablation) | MA-Mamba augmentations · MVSEC zero-shot | Abs. Rel: 0.35 -> 0.31 | Improved |
| increase STA receptive field from 8x8 to 32x32 (ablation) | MA-Mamba · MVSEC zero-shot | Abs. Rel: 0.37 -> 0.31 | Improved |
| use global STA receptive field instead of 32x32 (ablation) | MA-Mamba · MVSEC zero-shot | Abs. Rel: 0.31 -> 0.32 | Hurt |
| MobileMamba spatial extractor vs EfficientNet-V3 (ablation) | MA-Mamba · MVSEC zero-shot | Abs. Rel: 0.34 -> 0.31 | Improved |
| enlarge synthetic training set ~1.5x (ablation) | MA-Mamba · MVSEC zero-shot | Abs. Rel: 0.31 -> 0.27 | Improved |
| MA-Mamba vs best baseline RVT* | MA-Mamba · DSEC zero-shot | Abs. Rel: 0.29 -> 0.20 | Improved |
Closest papers by meaning
- EventMamba: Enhancing Spatio-Temporal Locality with State Space Models for Event-Based Video Reconstruction (AAAI 2025 · similarity 0.90)
- Event-Driven Dynamic Scene Depth Completion (NeurIPS 2025 · similarity 0.88)
- AIMDepth: Asymmetric Image-Event Mamba for Monocular Depth Estimation (CVPR 2026 · similarity 0.88)
- EventVGGT: Exploring Cross-Modal Distillation for Consistent Event-based Depth Estimation (ECCV 2026 · similarity 0.87)
- Talk2Event: Grounded Understanding of Dynamic Scenes from Event Cameras (NeurIPS 2025 · similarity 0.86)
- Motion-adaptive Transformer for Event-based Image Deblurring (AAAI 2025 · similarity 0.86)
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.
