Research map / Label-efficient and robust learning / Noisy labels
STARS: Spatial-Temporal Active Re-sampling for Label-Efficient Learning from Noisy Annotations
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 |
|---|---|---|---|
| Use STARS for active re-sampling instead of LOSS, random re-sampling, or no re-sampling | Active learning with noisy annotations · Yeast, Autodrive, Dermatology 1, Dermatology 2, Penstroke | accuracy | Helped |
| Use STARS instead of DEC or LIC alone (ablation) | Active learning with noisy annotations · Yeast, Autodrive, Dermatology 1, Dermatology 2, Penstroke | accuracy | Helped |
| Use STARS instead of LOSS or random re-sampling | Deep neural network active learning · FashionMNIST | accuracy | Helped |
| Use STARS instead of LOSS or random re-sampling | Deep neural network active learning · CIFAR10 | accuracy | Helped |
| Use loss-based re-sampling instead of random re-sampling (ablation) | Deep neural network active learning · FashionMNIST and CIFAR10 | accuracy | No clear effect |
| Replace the joint spatial-temporal criterion with DEC alone (ablation) | STARS · Yeast, Autodrive, Dermatology 1, Dermatology 2, Penstroke | accuracy | Hurt |
| Replace the joint spatial-temporal criterion with LIC alone (ablation) | STARS · Yeast, Autodrive, Dermatology 1, Dermatology 2, Penstroke | accuracy | Hurt |
| Use active re-sampling instead of no re-sampling | Active learning with noisy annotations · Synthetic data | accuracy | Helped |
| Use STARS instead of LOSS (ablation) | Active learning with noisy annotations · Synthetic data | accuracy | Helped |
| Correct on-margin noisy instances before outlier noisy instances (ablation) | Temporal re-sampling for maximum-margin classifiers · Synthetic data | accuracy | Helped |
| Use majority-vote re-sampling instead of an active re-sampling function | Active learning with noisy annotations · Real-world datasets | accuracy | Hurt |
Closest papers by meaning
- Re-Active Learning: Active Learning with Relabeling (AAAI 2016 · similarity 0.90)
- Boosting Active Learning via Improving Test Performance (AAAI 2022 · similarity 0.87)
- Towards Robust and Reproducible Active Learning Using Neural Networks (CVPR 2022 · similarity 0.87)
- Towards Cost-Effective Learning: A Synergy of Semi-Supervised and Active Learning (CVPR 2025 · similarity 0.87)
- A Cross-Domain Benchmark for Active Learning (NeurIPS 2024 · similarity 0.86)
- Enhancing Semi-Supervised Learning via Representative and Diverse Sample Selection (NeurIPS 2024 · similarity 0.86)
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.