Notice: This website is continuously updated as new conference proceedings and journal volumes are published. Consequently, the figures presented here may differ from those reported in our publications.
A brief overview of our methodology is provided on the Methods page, along with references to publications documenting our approach and links to the code and data used in each study.
The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates.
Beyond Pretrained Features: Noisy Image Modeling Provides Adversarial Defense
Authors: Zunzhi You, Daochang Liu, Bohyung Han, Chang Xu
NeurIPS 2023 | Venue PDF | LLM Run Details | Input Tokens: 16,384 Total number of tokens sent to the LLM as input for this paper's analysis. | Output Tokens: 4,794 Total number of tokens produced by the LLM (including reasoning/thinking tokens) for this paper's analysis.
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results demonstrate that, in terms of adversarial robustness, NIM is superior to MIM thanks to its effective denoising capability. In this section, we empirically show the effectiveness of NIM as a self-supervised pretraining method that can bring adversarial robustness via the De3 method. |
| Researcher Affiliation | Academia | Zunzhi You1, Daochang Liu1, Bohyung Han2, Chang Xu1 1 School of Computer Science, University of Sydney 2 ECE & IPAI, Seoul National University |
| Pseudocode | No | The paper does not contain structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | Source code and models are available at https://github.com/youzunzhi/NIM-AdvDef. |
| Open Datasets | Yes | We conduct all the experiments on Image Net-1K [11] dataset. [11] Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pages 248 255. Ieee, 2009. |
| Dataset Splits | Yes | We use the training set (1.28 million images) in pretraining and finetuning and use the validation set (50,000 images) for evaluation. |
| Hardware Specification | No | The paper does not provide specific details regarding the hardware used for running the experiments (e.g., GPU/CPU models, memory, or cloud instances). |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers (e.g., Python version, library versions like PyTorch or TensorFlow). |
| Experiment Setup | Yes | For PGD, we set the number of steps n = 10 and step size α = 2/255. Our default models are pretrained with σ Γ(25, 3) and finetuned on denoised images of σ U(0, 30). All models are pretrained for 800 epochs and then finetuned for 100 epochs. |