Research map / Label-efficient and robust learning / Noisy labels
Faster Meta Update Strategy for Noise-Robust Deep Learning
Check 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 |
|---|---|---|---|
| Apply FaMUS to MW-Net | MW-Net · CIFAR-10, 60% symmetric noise | accuracy: 84.5 -> 85.8 | Helped |
| Apply FaMUS to MW-Net | MW-Net · CIFAR-100, 60% symmetric noise | time per iteration: 933 -> 284 | Helped |
| Use FaMUS | robust learning with PreAct ResNet-18 · CIFAR-100, 60% symmetric noise | accuracy: 72.00 -> 73.58 | Helped |
| Use FaMUS | robust learning with Inception-ResNet V2 · WebVision validation | top-1 accuracy: 77.32 -> 79.40 | Helped |
| Use FaMUS | robust learning with PreAct ResNet-18 · CNWL, 60% noise | accuracy: 43.80 -> 45.10 | Helped |
| Add the layer-count regularizer L_r (ablation) | FaMUS on MW-Net · CIFAR-100, 60% symmetric noise | time per iteration: 933 -> 275 | Helped |
| Add the layer-count regularizer L_r (ablation) | FaMUS on MW-Net · CIFAR-100, 60% symmetric noise | accuracy: 61.7 -> 62.0 | Helped |
| Add the gradient-matching loss L_g (ablation) | FaMUS on MW-Net with L_r · CIFAR-100, 60% symmetric noise | accuracy: 62.0 -> 62.9 | Helped |
| Add the gradient-matching loss L_g (ablation) | FaMUS on MW-Net with L_r · CIFAR-100, 60% symmetric noise | time per iteration: 275 -> 284 | Hurt |
| Use a fixed residual block rather than learned layer sampling (ablation) | FaMUS on MW-Net · CIFAR-100, 60% symmetric noise | accuracy: 62.9 vs 60.8; 61.9; 61.4 | Hurt |
| Randomly sample 4 layers rather than learn which layers to sample (ablation) | FaMUS on MW-Net · CIFAR-100, 60% symmetric noise | accuracy: 61.2 vs 62.9 | Hurt |
| Vary the number of randomly sampled layers (ablation) | FaMUS on MW-Net · CIFAR-100, 60% symmetric noise | accuracy: 61.2 -> 61.6 -> 62.1 | Helped |
Closest papers by meaning
- Dynamic Kernel Selection for Improved Generalization and Memory Efficiency in Meta-Learning (CVPR 2022 · similarity 0.87)
- MetaDiff: Meta-Learning with Conditional Diffusion for Few-Shot Learning (AAAI 2024 · similarity 0.86)
- MetaOptimize: A Framework for Optimizing Step Sizes and Other Meta-parameters (ICML 2025 · similarity 0.85)
- Memory-Reduced Meta-Learning with Guaranteed Convergence (AAAI 2025 · similarity 0.85)
- MADA: Meta-Adaptive Optimizers Through Hyper-Gradient Descent (ICML 2024 · similarity 0.85)
- Provable Meta-Learning with Low-Rank Adaptations (NeurIPS 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.