Research map / Image generation / Autoregressive and latent generation
G-IR: Geometric Image Representation for Learning
Xin Chen, Qi Zhao, Wei Zeng, Zongben Xu
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 OT optimization and QC optimization to the direct G-IR pipeline (ablation) | G-IR · MNIST/CIFAR-10/CelebA-HQ restoration | PSNR: 22.020 -> 33.279 (MNIST); 24.644 -> 44.451 (CIFAR-10); 19.871 -> 32.173 (CelebA-HQ) | Improved |
| add OT optimization (vs w/o OTopt) (ablation) | G-IR · CelebA-HQ restoration | PSNR: 23.605 -> 32.173 | Improved |
| add QC optimization (vs w/o QCopt) (ablation) | G-IR · CelebA-HQ restoration | PSNR: 25.136 -> 32.173 | Improved |
| add OT optimization (vs w/o OTopt) (ablation) | G-IR · MNIST restoration | LPIPS: 0.201 -> 0.005 | Improved |
| train autoencoder on G-IR instead of pixel intensities | U-Net autoencoder · MNIST latent interpolation | adjacent LPIPS: 0.0280 -> 0.0156 (1-7); 0.0382 -> 0.0262 (1-0); 0.0130 -> 0.0109 (4-9) | Improved |
Closest papers by meaning
- Geometric Algebra Planes: Convex Implicit Neural Volumes (ICML 2025 · similarity 0.83)
- Discriminating image representations with principal distortions (ICLR 2025 · similarity 0.83)
- EigenGS Representation: From Eigenspace to Gaussian Image Space (CVPR 2025 · similarity 0.83)
- Structuring Representation Geometry with Rotationally Equivariant Contrastive Learning (ICLR 2024 · similarity 0.83)
- Learning Robust Generalizable Radiance Field with Visibility and Feature Augmented Point Representation (ICLR 2024 · similarity 0.82)
- GS-IR: 3D Gaussian Splatting for Inverse Rendering (CVPR 2024 · similarity 0.82)
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.
