Research map / Label-efficient and robust learning / Self-supervised and contrastive
solo-learn: A Library of Self-supervised Methods for Visual Representation Learning
Victor Guilherme Turrisi da Costa, Enrico Fini, Moin Nabi, Nicu Sebe, Elisa Ricci
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 Nvidia DALI data loading instead of PyTorch Vision | Barlow Twins · ImageNet-100 | training time (20 epochs): 1h 38m 27s -> 43m 2s | Improved |
| Use Nvidia DALI data loading instead of PyTorch Vision | BYOL · ImageNet-100 | training time (20 epochs): 1h 38m 46s -> 50m 33s | Improved |
| Use Nvidia DALI data loading instead of PyTorch Vision | NNCLR · ImageNet-100 | training time (20 epochs): 1h 38m 30s -> 42m 3s | Improved |
| Use DALI loading, accepting higher memory | NNCLR · ImageNet-100 | GPU memory: 5060 MB -> 9244 MB | Hurt |
| Use solo-learn implementation instead of Lightly | MoCo V2+ · CIFAR-10 | accuracy: 90.0 -> 92.94 | Improved |
| Use solo-learn implementation instead of Lightly | SimCLR · CIFAR-10 | accuracy: 89.0 -> 90.74 | Improved |
| Use solo-learn implementation instead of Lightly | SimSiam · CIFAR-10 | accuracy: 91.0 -> 90.51 | Hurt |
| Use online instead of offline linear evaluation | SimSiam · ImageNet-100 | top-1 accuracy: 78.72 -> 74.54 | Hurt |
| Use online instead of offline linear evaluation | Barlow Twins · ImageNet-100 | top-1 accuracy: 80.16 -> 80.38 | No clear effect |
Closest papers by meaning
- Harnessing small projectors and multiple views for efficient vision pretraining (NeurIPS 2024 · similarity 0.89)
- Towards Efficient and Effective Self-Supervised Learning of Visual Representations (ECCV 2022 · similarity 0.89)
- Revisiting Self-Supervised Visual Representation Learning (CVPR 2019 · similarity 0.89)
- Scaling and Benchmarking Self-Supervised Visual Representation Learning (ICCV 2019 · similarity 0.88)
- SEED: Self-supervised Distillation For Visual Representation (ICLR 2021 · similarity 0.87)
- Can Generative Models Improve Self-Supervised Representation Learning? (AAAI 2025 · similarity 0.87)
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.
