Research map / 3D vision / Face and shape
Deep Active Surface Models
Udaranga Wickramasinghe, Pascal Fua, Graham Knott
Check 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 DASM smoothing to Mesh R-CNN | Mesh R-CNN · ShapeNetCore.v1 test | Chamfer distance: 0.189 -> 0.183 | Improved |
| Add adaptive DASM smoothing to Mesh R-CNN | Mesh R-CNN · ShapeNetCore.v1 test | Normal distance: 0.691 -> 0.727 | Improved |
| Add adaptive DASM smoothing to Mesh R-CNN | Mesh R-CNN · ShapeNetCore.v1 test | F^0.3: 80.4 -> 81.0 | Improved |
| Add adaptive DASM smoothing to Voxel2Mesh | Voxel2Mesh λ=0.025 · CortexEM pre-synaptic and post-synaptic | IoU: 76.8 $\pm$ 0.9 -> 79.4 $\pm$1.4; 81.2 $\pm$1.4 -> 85.5 $\pm$ 1.2 | Improved |
| Add adaptive DASM smoothing to Voxel2Mesh | Voxel2Mesh λ=0.025 · CortexEM synapse | IoU: 65.4 $\pm$1.6 -> 65.3 $\pm$0.9 | No clear effect |
| Apply adaptive smoothing during model training rather than only as post-processing (ablation) | Mesh R-CNN · ShapeNetCore.v1 test | Chamfer distance: 0.249 -> 0.183 | Improved |
| Use adaptive instead of uniform smoothing (ablation) | DASM · ShapeNetCore.v1 test | Chamfer distance: 0.201 -> 0.183 | Improved |
| Add adaptive DASM smoothing when λ_edge=0.0 | Mesh R-CNN · ShapeNetCore.v1 test, λ_edge=0.0 | Chamfer distance: 0.144 -> 0.167 | Hurt |
| Increase the edge-length regularization weight to λ_edge=1.0 (ablation) | Mesh R-CNN versus Adaptive DASM · ShapeNetCore.v1 test | F^0.1: 29.7 vs 29.3 | No clear effect |
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.
- Pixel2Mesh — Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images (ECCV 2018) · better on 5 of 5 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
- SURGE: Surface Regularized Geometry Estimation from a Single Image (NeurIPS 2016 · similarity 0.85)
- Flatten Anything: Unsupervised Neural Surface Parameterization (NeurIPS 2024 · similarity 0.85)
- Deep Geometric Prior for Surface Reconstruction (CVPR 2019 · similarity 0.85)
- ParSeNet: A Parametric Surface Fitting Network for 3D Point Clouds (ECCV 2020 · similarity 0.85)
- SurfNet: Generating 3D Shape Surfaces Using Deep Residual Networks (CVPR 2017 · similarity 0.85)
- VNC: A Scale-Space Foundation for Learnable 3D Surface Evolution (ECCV 2026 · similarity 0.85)
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.
