Research map / Label-efficient and robust learning / Recognition and long-tail
Modeling Inter-Class and Intra-Class Constraints in Novel Class Discovery
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 |
|---|---|---|---|
| Add inter-class sKLD separating labelled and unlabelled examples (ablation) | UNO / IIC · CIFAR10 | ACC: 93.25 -> 99.07 | Helped |
| Add inter-class sKLD separating labelled and unlabelled examples (ablation) | UNO / IIC · CIFAR100-20 | ACC: 90.54 -> 92.17 | Helped |
| Add inter-class sKLD separating labelled and unlabelled examples (ablation) | UNO / IIC · CIFAR100-50 | ACC: 62.27 -> 64.31 | Helped |
| Add intra-class sKLD consistency between images and augmentations (ablation) | UNO / IIC · CIFAR10 | ACC: 93.25 -> 93.60 | No clear effect |
| Add intra-class sKLD consistency between images and augmentations (ablation) | UNO / IIC · CIFAR100-20 | ACC: 90.54 -> 91.13 | Helped |
| Add intra-class sKLD consistency between images and augmentations (ablation) | UNO / IIC · CIFAR100-50 | ACC: 62.27 -> 63.35 | Helped |
| Combine inter-class and intra-class sKLD constraints (ablation) | UNO / IIC · CIFAR100-20 | ACC: 92.17 -> 92.48 | Helped |
| Combine inter-class and intra-class sKLD constraints (ablation) | UNO / IIC · CIFAR100-50 | ACC: 64.31 -> 65.85 | Helped |
| Replace intra-class sKLD consistency with MSE (ablation) | UNO baseline · CIFAR10 | ACC: 93.25 -> 93.11 | Hurt |
| Replace intra-class sKLD consistency with MSE (ablation) | UNO baseline · CIFAR100-20 | ACC: 90.54 -> 90.96 | Helped |
| Replace intra-class sKLD consistency with MSE (ablation) | UNO baseline · CIFAR100-50 | ACC: 62.27 -> 62.47 | Helped |
| Use intra-class sKLD instead of MSE consistency (ablation) | UNO baseline · CIFAR10 | ACC: 93.25 -> 93.60 | Helped |
Closest papers by meaning
- Joint Class-level and Instance-level Relationship Modeling for Novel Class Discovery (AAAI 2025 · similarity 0.90)
- Adaptive Discovering and Merging for Incremental Novel Class Discovery (AAAI 2024 · similarity 0.86)
- Exploiting Inter-sample and Inter-feature Relations in Dataset Distillation (CVPR 2024 · similarity 0.85)
- Cross-domain Open-world Discovery (ICML 2024 · similarity 0.85)
- Novel Class Discovery in Chest X-rays via Paired Images and Text (AAAI 2024 · similarity 0.85)
- Unsupervised Learning for Class Distribution Mismatch (ICML 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.