Research map / Label-efficient and robust learning / Self-supervised and contrastive
Decoding Dynamic Visual Experience from Calcium Imaging via Cell-Pattern-Aware SSL
Sangyoon Bae, Mehdi Azabou, Blake Aaron Richards, Jiook Cha
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 |
|---|---|---|---|
| pretrain on predictable neurons then finetune on unpredictable | from-scratch POYO+ · movie decoding | SSIM: 0.528 -> 0.593 | Improved |
| pretrain on predictable neurons then finetune on unpredictable | from-scratch POYO+ · drifting gratings | accuracy: 0.492 -> 0.555 | Improved |
| reverse curriculum: pretrain on unpredictable neurons (ablation) | POYO-CAP · drifting gratings | accuracy: 0.555 -> 0.213 | Hurt |
| select neurons by statistical regularity instead of random subset (ablation) | POYO-CAP · drifting gratings | accuracy: 0.254 -> 0.555 | Improved |
| select predictable neurons instead of mixed with unpredictable (ablation) | POYO-CAP · movie decoding | SSIM: 0.543 -> 0.593 | Improved |
| use all predictable lines instead of inhibitory only (ablation) | POYO-CAP · movie decoding | SSIM: 0.544 -> 0.593 | Improved |
| temporal masking instead of random masking (ablation) | POYO-CAP · movie decoding | SSIM: 0.540 -> 0.593 | Improved |
| add auxiliary cross-entropy loss to masked reconstruction (ablation) | POYO-CAP · drifting gratings | accuracy: 0.099 -> 0.555 | Improved |
| add masked reconstruction to cross-entropy (ablation) | POYO-CAP · drifting gratings | accuracy: 0.452 -> 0.555 | Improved |
| CE weight 0.01 versus 0.1 (ablation) | POYO-CAP · movie decoding | SSIM: 0.552 -> 0.593 | Improved |
| add UNet skip connections (ablation) | POYO+ encoder with UNet decoder · movie decoding | SSIM: 0.466 -> 0.593 | Improved |
| pretrain on predictable neurons only | model scaling · SSIM vs model size | scaling slope: slope 0.018 vs approx 0.005-0.013 | Improved |
Closest papers by meaning
- Towards a "Universal Translator" for Neural Dynamics at Single-Cell, Single-Spike Resolution (NeurIPS 2024 · similarity 0.87)
- POCO: Scalable Neural Forecasting through Population Conditioning (NeurIPS 2025 · similarity 0.87)
- Leveraging unlabelled data for generalizable neural population decoding (NeurIPS 2026 · similarity 0.86)
- SpikeSSL: A Universal Spike Inference Framework with Dynamics-Informed State-Space Layers (NeurIPS 2026 · similarity 0.85)
- A Scalable, Causal, and Energy Efficient Framework for Neural Decoding with Spiking Neural Networks (NeurIPS 2025 · similarity 0.85)
- Generalizable, real-time neural decoding with hybrid state-space models (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.
