Research map / Image generation / Autoregressive and latent generation
Learning Conditional Deformable Templates with Convolutional Networks
Adrian Dalca, Marianne Rakic, John Guttag, Mert Sabuncu
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 |
|---|---|---|---|
| Use the learned unconditional template instead of the baseline atlas | Brain MRI registration and segmentation · Brain MRI test set | Dice: 0.731~(\pm0.153) -> 0.800~(\pm0.110) | Improved |
| Condition the learned template on age and sex rather than use an unconditional template | Learned brain MRI template · Brain MRI test set | Dice: 0.800~(\pm0.110) -> 0.795~(\pm0.116) | No clear effect |
| Add scale as a conditioning attribute (ablation) | Conditional template model · MNIST D-class-scale | Scale-related confounding in principal geometric variability | Improved |
| Replace deformation-based template learning with a decoder trained using pixel-wise MSE (ablation) | Template construction · MNIST D-class | Template centrality and deformation measures | Hurt |
| Leave some scale values or most examples of a class unobserved during training (ablation) | Conditional template model · MNIST D-class-scale | Quality of templates for held-out attributes | No clear effect |
| Use scale as an attribute rather than leave it unconditioned (ablation) | Geometric variability representation · MNIST D-class-scale | Scale dominance in principal components | Improved |
| Use deformation-minimizing representation instead of a single-neuron MSE autoencoder (ablation) | Latent-attribute image representation · MNIST D-class | Geometric template representation | Improved |
Closest papers by meaning
- Generative Adversarial Registration for Improved Conditional Deformable Templates (ICCV 2021 · similarity 0.91)
- ImplicitAtlas: Learning Deformable Shape Templates in Medical Imaging (CVPR 2022 · similarity 0.88)
- Anatomical Priors in Convolutional Networks for Unsupervised Biomedical Segmentation (CVPR 2018 · similarity 0.87)
- Anatomy-Preserving Unpaired Medical Image Translation via Shared Latent Anchoring (NeurIPS 2026 · similarity 0.87)
- An Unsupervised Learning Model for Deformable Medical Image Registration (CVPR 2018 · similarity 0.87)
- Representing Part-Whole Hierarchies in Foundation Models by Learning Localizability Composability and Decomposability from Anatomy via Self Supervision (CVPR 2024 · similarity 0.86)
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.
