Research map / Point clouds / Segmentation and detection
VaPiD: A Rapid Vanishing Point Detector via Learned Optimizers
Shichen Liu, Yichao Zhou, Yajie Zhao
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 efficient conic convolutions to the vanishing point proposal network (ablation) | VPPN · SU3 wireframe | recall at 6 deg: 89.53 -> 99.00 | Improved |
| add efficient conic convolutions to the vanishing point proposal network (ablation) | VPPN · SU3 wireframe | recall at 2 deg: 33.20 -> 56.20 | Improved |
| add efficient conic convolutions to the vanishing point proposal network (ablation) | VPPN · SU3 wireframe | mean angle error: 0.554° -> 0.146° | Improved |
| increase inference refinement steps from 4 to 8 (ablation) | NVPO · SU3 wireframe | median angle error: 0.145° -> 0.088° | Improved |
| increase inference refinement steps from 8 to 24 (ablation) | NVPO · SU3 wireframe | median angle error: 0.088° -> 0.085° | No clear effect |
| increase inference refinement steps from 4 to 24 (ablation) | NVPO · SU3 wireframe | FPS: 26 -> 7 | Hurt |
| increase inference refinement steps from 4 to 8 (ablation) | NVPO · SU3 wireframe | AA at 0.1 deg: 15.1 -> 26.6 | Improved |
| use VaPiD instead of NeurVPS | vanishing point detector · HoliCity | mean angle error: 8.32° -> 3.00° | Improved |
| use VaPiD instead of NeurVPS | vanishing point detector · Natural Scene | AA at 1 deg: 29.1 -> 24.6 | Hurt |
| use VaPiD instead of NeurVPS | vanishing point detector · Natural Scene | AA at 10 deg: 85.5 -> 87.4 | Improved |
| use VaPiD instead of CONSAC | vanishing point detector · NYU-VP | AUC at 10 deg: 65.0 -> 69.1 | Improved |
| use VaPiD instead of NeurVPS | vanishing point detector · SU3 wireframe | median angle error: 0.090° -> 0.088° | Improved |
Later papers that compared with this method
1 later accepted paper reports VaPiD as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Deep Vanishing Point Detection: Geometric Priors Make Dataset Variations Vanish (CVPR 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
- End-to-End Real-Time Vanishing Point Detection with Transformer (AAAI 2024 · similarity 0.89)
- Deep Vanishing Point Detection: Geometric Priors Make Dataset Variations Vanish (CVPR 2022 · similarity 0.89)
- VPDETR: End-to-End Vanishing Point DEtection TRansformers (AAAI 2024 · similarity 0.88)
- Convex Relaxation for Robust Vanishing Point Estimation in Manhattan World (CVPR 2025 · similarity 0.88)
- VGGT: Visual Geometry Grounded Transformer (CVPR 2025 · similarity 0.85)
- V-DETR: DETR with Vertex Relative Position Encoding for 3D Object Detection (ICLR 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.
