Research map / Label-efficient and robust learning / Recognition and long-tail
Contrastive Open Set Recognition
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 vanilla SupCon with SupCon-ST in ConOSR (ablation) | ConOSR · SVHN | AUROC: 98.8 -> 99.1 | Helped |
| Replace vanilla SupCon with SupCon-ST in ConOSR (ablation) | ConOSR · CIFAR-10 | AUROC: 93.7 -> 94.2 | Helped |
| Replace vanilla SupCon with SupCon-ST in ConOSR (ablation) | ConOSR · TinyImageNet | AUROC: 79.6 -> 80.9 | Helped |
| Use SupCon-ST instead of a plain CNN (ablation) | closed-set classifier · CIFAR-10 | accuracy: 94.0 -> 94.6 | Helped |
| Use SupCon-ST instead of a plain CNN (ablation) | closed-set classifier · CIFAR-100 | accuracy: 71.6 -> 73.0 | Helped |
| Use SupCon-ST instead of a plain CNN (ablation) | closed-set classifier · TinyImageNet | accuracy: 63.7 -> 66.1 | Helped |
| Train contrastive representations rather than plain CNN representations (ablation) | ConOSR · CIFAR-100 | AUROC: 76.7 -> 79.1 | Helped |
| Train contrastive representations rather than plain CNN representations (ablation) | ConOSR · TinyImageNet | AUROC: 68.1 -> 72.1 | Helped |
| Add mixup and label smoothing through SupCon-ST (ablation) | ConOSR · CIFAR-100 | accuracy on unknown classes: 62.6 -> 60.5 | Hurt |
| Use SupCon-ST instead of vanilla SupCon (ablation) | ConOSR · CIFAR-100 | AUROC: 77.9 -> 79.1 | Helped |
| Use ConOSR rather than PROSER | MNIST open-set recognition · MNIST + Omniglot unknown | macro F1: 86.2 -> 95.4 | Helped |
| Use ConOSR rather than PROSER | MNIST open-set recognition · MNIST + Noise unknown | macro F1: 88.2 -> 98.8 | Helped |
Closest papers by meaning
- All Beings Are Equal in Open Set Recognition (AAAI 2024 · similarity 0.88)
- Frequency Shuffling and Enhancement for Open Set Recognition (AAAI 2024 · similarity 0.87)
- Exploring Diverse Representations for Open Set Recognition (AAAI 2024 · similarity 0.87)
- From Coarse to Fine-Grained Open-Set Recognition (CVPR 2024 · similarity 0.87)
- Unlocking the Power of Open Set: A New Perspective for Open-Set Noisy Label Learning (AAAI 2024 · similarity 0.85)
- Unlocking Better Closed-Set Alignment Based on Neural Collapse for Open-Set Recognition (AAAI 2025 · 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.