Research map / Label-efficient and robust learning / Domain adaptation and generalization
BrainRiem: Riemannian Prototype Learning for Source-Free Cross-Site Brain Network Diagnosis
Kunyu Zhang, Tianxiang 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 Log-Euclidean mapping for SPD prototypes (ablation) | BrainRiem · REST-meta-MDD (LOSO) | accuracy: 5.7% drop when removed | Improved |
| add spectral calibration loss (Dirichlet energy) (ablation) | BrainRiem · REST-meta-MDD (LOSO) | accuracy | Improved |
| add manifold regularization loss (ablation) | BrainRiem · REST-meta-MDD (LOSO) | accuracy | Improved |
| add teacher-guided supervision and bi-level optimization (ablation) | BrainRiem · REST-meta-MDD (LOSO) | accuracy | Improved |
| use geometry-aware prototype initialization instead of Euclidean K-means (ablation) | BrainRiem · REST-meta-MDD (LOSO) | accuracy | Improved |
| use geometry-aware prototype initialization instead of random prototypes (ablation) | BrainRiem · REST-meta-MDD (LOSO) | accuracy: ~50% for random prototypes | Improved |
| increase prototypes per class K above 8 (ablation) | BrainRiem · ABIDE and REST-meta-MDD | accuracy | Hurt |
| set prototypes per class K=4 among {1, 2, 4, 8, 16} (ablation) | BrainRiem · ABIDE and REST-meta-MDD | accuracy | Improved |
| vary regularization weights lambda1-lambda4 within an order of magnitude (ablation) | BrainRiem · REST-meta-MDD (LOSO) | accuracy | No clear effect |
| transmit BrainRiem prototypes instead of raw FC matrices (re-identification) | BrainRiem · NYU privacy audit | Rank-1 identification rate: 100% -> 3.8% | Improved |
| transmit BrainRiem prototypes instead of raw FC matrices (membership inference) | BrainRiem · NYU privacy audit | attack AUC: 0.99 -> 0.53 | Improved |
| use BrainRiem instead of StruRW for LOSO adaptation | BrainRiem · REST-meta-MDD (LOSO) | accuracy: 64.0 -> 70.2 | Improved |
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.
- CORAL — CORAL: Disentangling Latent Representations in Long-Tailed Diffusion (NeurIPS 2025) · 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
- MAGNET: Manifold-Aware Graph Diffusion Network for Connectome Generation (NeurIPS 2026 · similarity 0.88)
- Riemannian Flow Matching for Brain Connectivity Matrices via Pullback Geometry (NeurIPS 2025 · similarity 0.88)
- Exploring the Enigma of Neural Dynamics Through A Scattering-Transform Mixer Landscape for Riemannian Manifold (ICML 2024 · similarity 0.86)
- SPDIM: Source-Free Unsupervised Conditional and Label Shift Adaptation in EEG (ICLR 2025 · similarity 0.85)
- Geodesic Optimization for Predictive Shift Adaptation on EEG data (NeurIPS 2024 · similarity 0.85)
- Brain Harmony: A Multimodal Foundation Model Unifying Morphology and Function into 1D Tokens (NeurIPS 2025 · 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.
