Research map / Point clouds / Shape and surface completion
GenSDF: Two-Stage Learning of Generalizable Signed Distance Functions
Gene Chou, Ilya Chugunov, Felix Heide
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 semi-supervised stage on top of meta-learning stage (ablation) | GenSDF · Acronym unseen classes | mean Chamfer Distance: 0.515 -> 0.407 | Improved |
| add meta-learning pretraining before semi-supervised stage (ablation) | GenSDF · Acronym unseen classes | mean Chamfer Distance: 0.889 -> 0.407 | Improved |
| add meta-learning stage on top of plain supervised training on 20 classes (ablation) | supervised baseline · Acronym unseen classes | mean Chamfer Distance: 1.164 -> 0.515 | Improved |
| episodic meta-learning vs supervised training on 76 classes (ablation) | supervised baseline · Acronym unseen classes | mean Chamfer Distance: 0.908 -> 0.515 | Improved |
| meta-learning vs supervised on seen classes (ablation) | supervised baseline · Acronym seen classes | mean Chamfer Distance: 0.695 -> 0.503 | Improved |
| add semi-supervised stage on seen classes (ablation) | GenSDF · Acronym seen classes | mean Chamfer Distance: 0.503 -> 0.351 | Improved |
| use signed nearest-neighbor self-supervised loss vs SAL/NeuralPull | unsupervised SDF learning · Four classes | mean Chamfer Distance: 2.419 -> 0.218 | Improved |
| add two-stage semi-supervised training over self-supervised only (ablation) | GenSDF · Four classes | mean Chamfer Distance: 0.218 -> 0.089 | Improved |
| increase Gaussian noise variance on input point cloud | GenSDF · YCB Pitcher | Chamfer Distance: 0.870 -> 5.208 (sigma^2 0 -> 0.1) | Hurt |
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.
- NeuralPull — Neural-Pull: Learning Signed Distance Function from Point clouds by Learning to Pull Space onto Surface (ICML 2021) · better on 2 of 2 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
- Few-Shot Unsupervised Implicit Neural Shape Representation Learning with Spatial Adversaries (ICML 2024 · similarity 0.89)
- DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation (CVPR 2019 · similarity 0.89)
- Inferring Neural Signed Distance Functions by Overfitting on Single Noisy Point Clouds through Finetuning Data-Driven based Priors (NeurIPS 2024 · similarity 0.89)
- Unsupervised Inference of Signed Distance Functions From Single Sparse Point Clouds Without Learning Priors (CVPR 2023 · similarity 0.88)
- SDF-SRN: Learning Signed Distance 3D Object Reconstruction from Static Images (NeurIPS 2020 · similarity 0.88)
- Small Steps and Level Sets: Fitting Neural Surface Models with Point Guidance (CVPR 2024 · similarity 0.87)
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.
