Research map / 3D vision / Depth and stereo
S2M2: Scalable Stereo Matching Model for Reliable Depth Estimation
Junhong Min, Youngpil Jeon, Jimin Kim, Minyong 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 Adaptive Gated Fusion Layer (ablation) | S2M2 Ch128 Ntr1 · SceneFlow | Bad-2.0: 1.974 -> 1.601 | Improved |
| add Adaptive Gated Fusion Layer (ablation) | S2M2 Ch128 Ntr1 · SceneFlow | EPE: 0.489 -> 0.496 | No clear effect |
| add PMC loss (ablation) | S2M2 Ch128 Ntr1 · SceneFlow | EPE: 0.496 -> 0.451 | Improved |
| add PMC loss (ablation) | S2M2 Ch128 Ntr1 · SceneFlow | Bad-2.0: 1.601 -> 1.490 | Improved |
| add PMC loss (ablation) | S2M2 Ch256 Ntr3 · SceneFlow | EPE: 0.299 -> 0.254 | Improved |
| add PMC loss (ablation) | S2M2 Ch256 Ntr3 · SceneFlow | Bad-2.0: 0.810 -> 0.750 | Improved |
| increase transformer blocks from 1 to 3 (ablation) | S2M2 Ch128 · SceneFlow | EPE: 0.451 -> 0.388 | Improved |
| increase channels from 128 to 256 (ablation) | S2M2 Ntr1 · SceneFlow | Bad-2.0: 1.490 -> 0.928 | Improved |
| add local refinement iterations (0 to 3) (ablation) | S2M2 Ch128 Ntr1 · SceneFlow | EPE: 0.666 -> 0.451 | Improved |
| add local refinement iterations (0 to 3) (ablation) | S2M2 Ch128 Ntr1 · SceneFlow | GFLOPs: 305 -> 647 | Hurt |
| add a second refinement iteration (1 to 3) (ablation) | S2M2 Ch128 Ntr1 · SceneFlow | EPE: 0.470 -> 0.451 | No clear effect |
| fine-tune on KITTI (ablation) | Sel-IGEV · KITTI 2015 D1-all non-occ | D1-all: 4.57 -> 1.44 | Improved |
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.
- IGEV — IGEV++: Iterative Multi-Range Geometry Encoding Volumes for Stereo Matching (TPAMI 2025) · 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
- Revisiting Stereo Depth Estimation From a Sequence-to-Sequence Perspective With Transformers (ICCV 2021 · similarity 0.92)
- Single View Stereo Matching (CVPR 2018 · similarity 0.91)
- Stereo Anywhere: Robust Zero-Shot Deep Stereo Matching Even Where Either Stereo or Mono Fail (CVPR 2025 · similarity 0.90)
- Any-Stereo: Arbitrary Scale Disparity Estimation for Iterative Stereo Matching (AAAI 2024 · similarity 0.90)
- MonSter: Marry Monodepth to Stereo Unleashes Power (CVPR 2025 · similarity 0.89)
- PPMStereo: Pick-and-Play Memory Construction for Consistent Dynamic Stereo Matching (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.
