Research map / Label-efficient and robust learning / Self-supervised and contrastive
FroSSL: Frobenius Norm Minimization for Efficient Multiview Self-Supervised Learning
Oscar Skean, Aayush Dhakal, Nathan Jacobs, Luis G Sanchez Giraldo
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 |
|---|---|---|---|
| increase FroSSL views from 2 to 8 (ablation) | FroSSL · STL-10 linear probe | top-1 accuracy: 87.3 -> 90.9 | Improved |
| increase FroSSL views from 2 to 8 (ablation) | FroSSL · Tiny ImageNet linear probe | top-1 accuracy: 44.2 -> 45.3 | Improved |
| increase FroSSL views from 2 to 8 (ablation) | FroSSL · CIFAR-10/100 | top-1 accuracy | No clear effect |
| increase FroSSL views from 2 to 8 (ablation) | FroSSL · ImageNet-1k 10% data | top-1 accuracy: 33.4 -> 38.2 | Improved |
| add log to Frobenius norm term (ablation) | FroSSL (2 views) · Average over 5 datasets | accuracy: 69.7 -> 74.6 | Improved |
| use strong instead of weak augmentations (ablation) | FroSSL (2 views) · Tiny ImageNet | top-1 accuracy: 39.4 -> 44.2 | Improved |
| FroSSL vs other dimension-contrastive methods under weak augmentations | FroSSL · Tiny ImageNet | top-1 accuracy: VICReg 18.1, MMCR 24.2 vs FroSSL 39.4 | Improved |
| use FroSSL (2 views) instead of other SSL methods | FroSSL · Average over 5 datasets | accuracy: 74.7 (Barlow Twins) vs 74.6 | No clear effect |
| use FroSSL in low data regime | FroSSL · ImageNet-1k 10% data | top-1 accuracy: 31.1 (SimCLR) -> 33.4 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Barlow Twins — Barlow Twins: Self-Supervised Learning via Redundancy Reduction (ICML 2021) · better on 2 of 3 shared results
- VICReg — VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learning (ICLR 2022) · better on 1 of 1 shared result
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
- Self-Supervised Learning from Structural Invariance (ICLR 2026 · similarity 0.87)
- Preventing Dimensional Collapse in Self-Supervised Learning via Orthogonality Regularization (NeurIPS 2024 · similarity 0.85)
- Harnessing small projectors and multiple views for efficient vision pretraining (NeurIPS 2024 · similarity 0.85)
- Multi-View Causal Representation Learning with Partial Observability (ICLR 2024 · similarity 0.84)
- Collapse-Proof Non-Contrastive Self-Supervised Learning (ICML 2025 · similarity 0.84)
- Latent Functional Maps: a spectral framework for representation alignment (NeurIPS 2024 · similarity 0.84)
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.
