Research map / Point clouds / Shape and surface completion
Implicit Surface Representations As Layers in Neural Networks
Mateusz Michalkiewicz, Jhony K. Pontes, Dominic Jack, Mahsa Baktashmotlagh, Anders Eriksson
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 |
|---|---|---|---|
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Bottle (20^3 resolution) | IoU (%): 65.7 -> 78.4 (+12.7) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Car (20^3 resolution) | IoU (%): 73.0 -> 86.6 (+13.6) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Chair (20^3 resolution) | IoU (%): 57.2 -> 63.7 (+6.5) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Sofa (20^3 resolution) | IoU (%): 62.1 -> 68.9 (+6.8) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Telephone (20^3 resolution) | IoU (%): 60.5 -> 73.5 (+13.0) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Bottle (32^3 resolution) | IoU (%): 71.9 -> 78.4 (+6.5) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Car (32^3 resolution) | IoU (%): 81.9 -> 86.0 (+4.1) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Chair (32^3 resolution) | IoU (%): 59.2 -> 61.9 (+2.7) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Sofa (32^3 resolution) | IoU (%): 68.0 -> 72.9 (+4.9) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Telephone (32^3 resolution) | IoU (%): 67.8 -> 71.9 (+4.1) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Car (20^3 resolution) | Chamfer distance: 0.0917 -> 0.0411 (-0.0506) | Improved |
| replace voxel occupancy output with level set representation | 3D auto-encoder with single-image encoder · ShapeNet Bottle (20^3 resolution) | Chamfer distance: 0.0895 -> 0.0593 (-0.0302) | Improved |
Closest papers by meaning
- Learning to Infer Implicit Surfaces without 3D Supervision (NeurIPS 2019 · similarity 0.91)
- PatchNets: Patch-Based Generalizable Deep Implicit 3D Shape Representations (ECCV 2020 · similarity 0.91)
- DeepCurrents: Learning Implicit Representations of Shapes With Boundaries (CVPR 2022 · similarity 0.90)
- Implicit Geometric Regularization for Learning Shapes (ICML 2020 · similarity 0.90)
- Iso-Points: Optimizing Neural Implicit Surfaces With Hybrid Representations (CVPR 2021 · similarity 0.90)
- LISR: Learning Linear 3D Implicit Surface Representation Using Compactly Supported Radial Basis Functions (AAAI 2024 · similarity 0.88)
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.
