Research map / Adversarial robustness and security / Black-box and transfer attacks
REST: Performance Improvement of a Black Box Model via RL-Based Spatial Transformation
Jae Myung Kim, Hyungjin Kim, Chanwoo Park, Jungwoo Lee
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 REST before the black-box classifier | MNIST black-box classifier · MNIST R | accuracy: 69.28 -> 97.63 | Improved |
| Add REST before the black-box classifier | MNIST black-box classifier · MNIST RSc | accuracy: 65.52 -> 96.30 | Improved |
| Add REST before the black-box classifier | MNIST black-box classifier · MNIST RSh | accuracy: 56.52 -> 94.81 | Improved |
| Add REST before the black-box classifier | MNIST black-box classifier · MNIST RSS | accuracy: 53.95 -> 93.20 | Improved |
| Add REST before the black-box classifier | MNIST black-box classifier · MNIST RSST | accuracy: 17.99 -> 85.05 | Improved |
| Add REST before the black-box classifier | SVHN classifier · SVHN RSST | accuracy: 26.09 -> 83.82 | Improved |
| Add REST before the black-box classifier | CIFAR10 classifier · CIFAR10 RSST | accuracy: 32.09 -> 60.27 | Improved |
| Add REST before the black-box classifier | STL10 classifier · STL10 RSST | accuracy: 30.90 -> 53.20 | Improved |
| Train with fewer RST transformation combinations | REST · MNIST unseen RST transformation compositions | accuracy | Improved |
| Train REST on 1000 affine-warped examples | REST · MNIST R, RSc, RSh, RSS, and RSST | accuracy | Improved |
Closest papers by meaning
- The Principle of Diversity: Training Stronger Vision Transformers Calls for Reducing All Levels of Redundancy (CVPR 2022 · similarity 0.85)
- CNN Filter DB: An Empirical Investigation of Trained Convolutional Filters (CVPR 2022 · similarity 0.85)
- Depth CNNs for RGB-D Scene Recognition: Learning from Scratch Better than Transferring from RGB-CNNs (AAAI 2017 · similarity 0.85)
- Tilt your Head: Activating the Hidden Spatial-Invariance of Classifiers (ICML 2024 · similarity 0.84)
- B-cosification: Transforming Deep Neural Networks to be Inherently Interpretable (NeurIPS 2024 · similarity 0.84)
- Learning Where to Edit Vision Transformers (NeurIPS 2024 · 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.
