Research map / Point clouds / Shape and surface completion
GAL: Geometric Adversarial Loss for Single-View 3D-Object Reconstruction
Li Jiang, Shaoshuai Shi, Xiaojuan Qi, Jiaya Jia
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 multi-view geometric loss to the baseline (ablation) | PointSetGeneration · ShapeNet (mean) | Chamfer distance x1e-4: 40.90 -> 39.62 | Improved |
| add multi-view geometric loss to the baseline (ablation) | PointSetGeneration · ShapeNet (mean) | voxel IoU (%): 68.07 -> 69.10 | Improved |
| add conditional adversarial loss to the baseline (ablation) | PointSetGeneration · ShapeNet (mean) | Chamfer distance x1e-4: 40.90 -> 39.79 | Improved |
| add conditional adversarial loss to the baseline (ablation) | PointSetGeneration · ShapeNet (mean) | voxel IoU (%): 68.07 -> 69.64 | Improved |
| combine geometric and adversarial losses (full GAL) (ablation) | PointSetGeneration · ShapeNet (mean) | Chamfer distance x1e-4: 40.90 -> 37.26 | Improved |
| combine geometric and adversarial losses (full GAL) (ablation) | PointSetGeneration · ShapeNet (mean) | voxel IoU (%): 68.07 -> 71.16 | Improved |
| add geometric loss to the baseline (ablation) | PointSetGeneration · ShapeNet 2D projection (192x256, mean) | 2D projection IoU: 0.332 -> 0.337 | Improved |
| add full GAL to the baseline (ablation) | PointSetGeneration · ShapeNet 2D projection (192x256, mean) | 2D projection IoU: 0.332 -> 0.344 | Improved |
| add full GAL to the baseline (ablation) | PointSetGeneration · ShapeNet 2D projection (96x128, mean) | 2D projection IoU: 0.613 -> 0.634 | Improved |
| add full GAL to the baseline (ablation) | PointSetGeneration · ShapeNet 2D projection (48x64, mean) | 2D projection IoU: 0.782 -> 0.801 | Improved |
| add conditional adversarial loss to the baseline, speaker category (ablation) | PointSetGeneration · ShapeNet (speaker) | Chamfer distance x1e-4: 44.10 -> 47.24 | Hurt |
| add conditional adversarial loss to the baseline, lamp category (ablation) | PointSetGeneration · ShapeNet (lamp) | Chamfer distance x1e-4: 39.19 -> 42.39 | Hurt |
Closest papers by meaning
- Manifold Adversarial Learning for Cross-Domain 3D Shape Representation (ECCV 2022 · similarity 0.88)
- Learning To Measure the Point Cloud Reconstruction Loss in a Representation Space (CVPR 2023 · similarity 0.88)
- MVPNet: Multi-View Point Regression Networks for 3D Object Reconstruction from A Single Image (AAAI 2019 · similarity 0.88)
- A Point Set Generation Network for 3D Object Reconstruction From a Single Image (CVPR 2017 · similarity 0.88)
- Hide in Thicket: Generating Imperceptible and Rational Adversarial Perturbations on 3D Point Clouds (CVPR 2024 · similarity 0.86)
- Explicitly Perceiving and Preserving the Local Geometric Structures for 3D Point Cloud Attack (AAAI 2024 · 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.
