Research map / 3D vision / Depth and stereo
Toward Ultrafast Depth Sensing via Active Event-Based Stereo Vision
Jianing Li, Yunjian Zhang, Haiqian Han, Kangyao Huang, 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 dynamic-interaction cost volume to baseline (ablation) | ActiveEventNet+ · Active-Event-Stereo (real) | EPE: 2.124 -> 1.968 | Improved |
| add dynamic-interaction cost volume on simulated data (ablation) | ActiveEventNet+ · RealSense-Event-Sim | EPE: 1.468 -> 1.209 | Improved |
| add MobileNet blocks to baseline (ablation) | ActiveEventNet+ · Active-Event-Stereo (real) | EPE: 2.124 -> 2.210 | Hurt |
| add ETC to MobileNet + cost volume (ablation) | ActiveEventNet+ · Active-Event-Stereo (real) | EPE: 1.993 -> 1.849 | Improved |
| add ETC to MobileNet + cost volume (simulated) (ablation) | ActiveEventNet+ · RealSense-Event-Sim | EPE: 1.223 -> 1.077 | Improved |
| use dynamic interaction instead of concatenation (ablation) | ActiveEventNet+ · Active-Event-Stereo (real) | EPE: 2.015 -> 1.849 | Improved |
| use dynamic interaction instead of attention-based fusion (ablation) | ActiveEventNet+ · Active-Event-Stereo (real) | EPE: 1.879 -> 1.849 | Improved |
| use ETC instead of ConvLSTM (ablation) | ActiveEventNet+ · Active-Event-Stereo (real) | D1-all: 0.089 -> 0.073 | Improved |
| use reconstructed images instead of event images (ablation) | ActiveEventNet+ · Active-Event-Stereo (real) | EPE: 1.849 -> 1.791 | Mixed |
| increase temporal aggregation length from 1 to 3 (ablation) | ActiveEventNet+ · Active-Event-Stereo (real) | EPE: 1.995 -> 1.849 | Improved |
| increase temporal aggregation length from 3 to 6 (ablation) | ActiveEventNet+ · Active-Event-Stereo (real) | EPE: 1.849 -> 1.782 | Mixed |
| use MobileNet v2 for both feature extraction and encoder-decoder instead of standard conv (ablation) | ActiveEventNet+ · Active-Event-Stereo (real) | EPE: 1.752 -> 1.849 | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PSMNet — PSMNet: Position-Aware Stereo Merging Network for Room Layout Estimation (CVPR 2022) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Active Event-based Stereo Vision (CVPR 2025 · similarity 0.98)
- Temporal Event Stereo via Joint Learning with Stereoscopic Flow (ECCV 2024 · similarity 0.92)
- A Low Power, High Throughput, Fully Event-Based Stereo System (CVPR 2018 · similarity 0.91)
- Temporal Stereo Matching From Event Cameras via Joint Learning With Stereoscopic Flow (TPAMI 2025 · similarity 0.91)
- Zero-Shot Event-Intensity Asymmetric Stereo via Visual Prompting from Image Domain (NeurIPS 2024 · similarity 0.88)
- DERD-Net: Learning Depth from Event-based Ray Densities (NeurIPS 2025 · 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.
