Research map / 3D vision / Depth and stereo
EGformer: Equirectangular Geometry-biased Transformer for 360 Depth Estimation
Ilwi Yun, Chanyong Shin, Hyunku Lee, Hyuk-Jae Lee, Chae Eun Rhee
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 |
|---|---|---|---|
| use EGformer instead of Panoformer | 360 depth estimation · Structured3D test | Abs.rel: 0.0394 -> 0.0342 | Improved |
| use EGformer instead of Panoformer | 360 depth estimation · Pano3D test | Abs.rel: 0.0699 -> 0.0660 | Improved |
| mix Panoformer-style and EH(V)-MSA blocks (MMEE-EEMM) instead of EH(V)-MSA only (ablation) | EGformer · Structured3D test | Abs.rel: 0.0375 -> 0.0342 | Improved |
| use horizontal-only attention instead of combined EH(V)-MSA (ablation) | EGformer · Structured3D test | Abs.rel: 0.0375 -> 0.0421 | Hurt |
| add Das (distance-aware score) (ablation) | EGformer · Structured3D test | Abs.rel: 0.0360 -> 0.0342 | Improved |
| add EaAR (ablation) | EGformer · Structured3D test | Abs.rel: 0.0363 -> 0.0342 | Improved |
| add ERPE, Das and EaAR over softmax attention with LePE (ablation) | EGformer · Pano3D test | Abs.rel: 0.0700 -> 0.0660 | Improved |
| use EH(V)-MSA instead of CSwin-style softmax+LePE attention in EEEE-E-EEEE (ablation) | EGformer · Structured3D test | Abs.rel: 0.0399 -> 0.0375 | Improved |
| set geometry bias level rho to 0.1 instead of 0.03 or 0.3 (ablation) | EGformer · Structured3D (trained on S3D only) | Abs.rel: 0.0347 / 0.0352 -> 0.0338 | Improved |
Later papers that compared with this method
1 later accepted paper reports EGFormer as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- PanoVGGT: Feed-Forward 3D Reconstruction from Panoramic Imagery (CVPR 2026 · better on 2 of 3 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Panoformer — PanoFormer: Panorama Transformer for Indoor 360° Depth Estimation (ECCV 2022) · better on 7 of 7 shared results
- Bifuse — BiFuse++: Self-Supervised and Efficient Bi-Projection Fusion for $360^{\circ }$ Depth Estimation (TPAMI 2022) · better on 1 of 1 shared result
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
- Improving 360 Monocular Depth Estimation via Non-local Dense Prediction Transformer and Joint Supervised and Self-Supervised Learning (AAAI 2022 · similarity 0.88)
- SphereUFormer: A U-Shaped Transformer for Spherical 360 Perception (CVPR 2025 · similarity 0.85)
- Elite360D: Towards Efficient 360 Depth Estimation via Semantic- and Distance-Aware Bi-Projection Fusion (CVPR 2024 · similarity 0.85)
- RRT-MVS: Recurrent Regularization Transformer for Multi-View Stereo (AAAI 2025 · similarity 0.84)
- HELVIPAD: A Real-World Dataset for Omnidirectional Stereo Depth Estimation (CVPR 2025 · similarity 0.84)
- Improving Depth Completion via Depth Feature Upsampling (CVPR 2024 · similarity 0.84)
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.
