Research map / Image generation / Autoregressive and latent generation
GuidedMixup: An Efficient Mixup Strategy Guided by Saliency Maps
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 Guided-AP instead of PuzzleMix | PreActResNet-18 · CIFAR-100 | Top-1 error (%): 19.62 -> 18.80 | Helped |
| Use Guided-AP instead of Co-Mixup | PreActResNet-18 · Tiny-ImageNet | Top-1 error (%): 35.85 -> 35.37 | Helped |
| Use Guided-AP instead of strongest listed baseline | ResNet-18 · CUB, Cars, Aircraft, and Caltech-101 | Top-1 accuracy (%) | Helped |
| Add AugMix to Guided-SR (ablation) | WideResNet28-10 · CIFAR-100-C | Corruption error (%): 42.89 -> 29.54 | Helped |
| Add AugMix to Guided-AP (ablation) | WideResNet28-10 · CIFAR-100-C | Corruption error (%): 41.11 -> 29.07 | Helped |
| Add the proposed pairing algorithm to CutMix (ablation) | CutMix · Tiny-ImageNet | Top-1 accuracy (%): +1.14 | Helped |
| Add the proposed pairing algorithm to SaliencyMix (ablation) | SaliencyMix · CUB | Top-1 accuracy (%): +0.40 | Helped |
| Use greedy saliency pairing rather than random pairing (ablation) | Guided-SR · CUB | Top-1 accuracy (%): +1.15 | Helped |
| Use greedy saliency pairing rather than random pairing (ablation) | Guided-AP · Aircraft | Top-1 accuracy (%): +0.19 | Helped |
| Use precise saliency detection (AP) instead of spectral residual saliency (SR) (ablation) | GuidedMixup · CUB, Cars, Aircraft, and Caltech-101 | Top-1 accuracy (%) | Helped |
Closest papers by meaning
- SageMix: Saliency-Guided Mixup for Point Clouds (NeurIPS 2022 · similarity 0.88)
- SaliencyMix: A Saliency Guided Data Augmentation Strategy for Better Regularization (ICLR 2021 · similarity 0.88)
- Catch-Up Mix: Catch-Up Class for Struggling Filters in CNN (AAAI 2024 · similarity 0.87)
- Adversarial AutoMixup (ICLR 2024 · similarity 0.87)
- ADD: Attribution-Driven Data Augmentation Framework for Boosting Image Super-Resolution (CVPR 2025 · similarity 0.85)
- TdAttenMix: Top-Down Attention Guided Mixup (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.