Research map / Learning theory and algorithms / Optimal transport
Optimal Transport for Brain-Image Alignment: Unveiling Redundancy and Synergy in Neural Information Processing
Yang Xiao, Wang Lu, Jie Ji, Ruimeng Ye, Gen Li, Xiaolong Ma, Bo Hui
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 MSE alignment loss with OT loss (ablation) | UMBRAE (single subject) · NSD brain captioning | CIDEr: 51.50 -> 71.52 | Improved |
| replace MSE alignment loss with OT loss (ablation) | UMBRAE (cross subjects) · NSD brain captioning | CIDEr: 58.8 -> 72.90 | Improved |
| replace MSE alignment loss with OT loss (ablation) | UMBRAE (single subject) · NSD brain captioning | average over metrics: 6.28% | Improved |
| use OT alignment instead of best prior method (UMBRAE-S1) | UMBRAE-S1 · NSD brain captioning | average over metrics: 6.11% | Improved |
| use OT alignment instead of UMBRAE (cross subjects) | UMBRAE · NSD brain captioning | average over metrics: 3.81% | Improved |
| train on cross subjects instead of single subject | OT alignment · NSD brain captioning | CIDEr: 71.52 -> 72.90 | Improved |
| train on cross subjects instead of single subject | OT alignment · NSD brain captioning | BLEU4: 21.72 -> 21.50 | No clear effect |
| use all five regions instead of vision only (ablation) | OT alignment · NSD brain captioning | CIDEr: 10.95 -> 62.83 | Improved |
| use all five regions instead of Place only (ablation) | OT alignment · NSD brain captioning | CIDEr: 23.00 -> 62.83 | Improved |
| add vision region to P+F+W+B (ablation) | OT alignment · NSD brain captioning | CIDEr: 55.59 -> 62.83 | Improved |
| add remaining 'other' voxels to the five regions (ablation) | OT alignment · NSD brain captioning | CIDEr: 62.83 -> 71.52 | Improved |
| use PCA/UMAP multi-region embeddings (V-PFBW) instead of 5 regions (ablation) | embedding visualization · NSD embedding visualization (UMAP) | S-Score: -0.0784 -> 0.1344 | Improved |
Closest papers by meaning
- Human-Aligned Image Models Improve Visual Decoding from the Brain (ICML 2025 · similarity 0.88)
- Multigranular Evaluation for Brain Visual Decoding (AAAI 2026 · similarity 0.87)
- MindAlign: Bridging EEG, Vision, and Language for Zero-Shot Visual Decoding (NeurIPS 2026 · similarity 0.87)
- Correlating instruction-tuning (in multimodal models) with vision-language processing (in the brain) (ICLR 2025 · similarity 0.87)
- Wills Aligner: Multi-Subject Collaborative Brain Visual Decoding (AAAI 2025 · similarity 0.86)
- Bridging Critical Gaps in Convergent Learning: How Representational Alignment Evolves Across Layers, Training, and Distribution Shifts (NeurIPS 2025 · 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.
