Research map / Label-efficient and robust learning / Recognition and long-tail
Learning Open Set Network with Discriminative Reciprocal Points
Guangyao Chen, Limeng Qiao, Yemin Shi, Peixi Peng, Jia Li, Tiejun Huang, Shiliang Pu, Yonghong Tian
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 |
|---|---|---|---|
| add open space risk term L_o (lambda 0 -> 0.1) (ablation) | RPL · Air-300 | AUROC: 64.6 -> 83.1 | Improved |
| add open space risk term L_o (lambda 0 -> 0.1) (ablation) | RPL · Air-300 | accuracy: 84.6 -> 88.8 | Improved |
| add open space risk term L_o (lambda 0 -> 0.1) (ablation) | RPL · ImageNet-LT | AUROC: 55.0 -> 55.1 | Improved |
| increase lambda from 0.1 to 1 (ablation) | RPL · Air-300 | accuracy: 88.8 -> 86.7 | Hurt |
| increase lambda from 0.1 to 1 (ablation) | RPL · Air-300 | AUROC: 83.1 -> 81.7 | Hurt |
| increase lambda from 0.01 to 0.1 (ablation) | RPL · Air-300 | AUPR-Unknown: 68.9 -> 70.2 | Improved |
| add prototype loss to RPL (RPL++) | RPL · Air-300 | AUROC: 83.1 -> 84.1 | Improved |
| add prototype loss to RPL (RPL++) | RPL · ImageNet-LT | accuracy: 39.0 -> 39.7 | Improved |
| add prototype loss to RPL (RPL++) | RPL · CIFAR10 open set identification | AUROC: 82.7 -> 86.1 | Improved |
| replace softmax with RPL | Softmax · Air-300 | AUROC: 77.1 -> 83.1 | Improved |
| replace softmax with RPL | Softmax · ImageNet-LT | accuracy: 37.8 -> 39.0 | Improved |
| replace GCPL with RPL | GCPL · Air-300 | AUROC: 79.1 -> 83.1 | Improved |
Closest papers by meaning
- Adversarial Reciprocal Points Learning for Open Set Recognition (TPAMI 2021 · similarity 0.93)
- Classifier-induced Reciprocal Points for Multi-label Open-set Recognition (AAAI 2026 · similarity 0.86)
- Few-Shot Open-Set Recognition Using Meta-Learning (CVPR 2020 · similarity 0.86)
- All Beings Are Equal in Open Set Recognition (AAAI 2024 · similarity 0.85)
- Exploring Diverse Representations for Open Set Recognition (AAAI 2024 · similarity 0.83)
- Unlocking Better Closed-Set Alignment Based on Neural Collapse for Open-Set Recognition (AAAI 2025 · similarity 0.83)
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.
