Vision Transformer Neural Architecture Search for Out-of-Distribution Generalization: Benchmark and Insights
Paper • 2501.03782 • Published
Datasets:
ArXiv:
License:
Dataset Viewer
The dataset viewer is not available for this split.
The number of columns (1001) exceeds the maximum supported number of columns (1000). This is a current limitation of the datasets viewer. You can reduce the number of columns if you want the viewer to work.
Error code: TooManyColumnsError
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Official dataset of "Vision Transformer Neural Architecture Search for Out-of-Distribution Generalization: Benchmark and Insights" Sy-Tuyen Ho*, Tuan Van Vo*, Somayeh Ebrahimkhani*, Ngai-Man Cheung. (* joint first authors) NeurIPS'24 (Main Track)
The dataset is divided into three AutoFormer search spaces: Supernet-Base, Supernet-Small, and Supernet-Tiny. Each dataset includes evaluation results (ID Accuracy, OoD Accuracy) for 8 large common OoD datasets: ImageNet-C, ImageNet-A, ImageNet-O, ImageNet-P, ImageNet-D, ImageNet-R, ImageNet-Sketch, and Stylized ImageNet. This structure helps keep file sizes manageable and allows for selective evaluation results.
Links
- Paper: arXiv:2501.03782
- Code: OoD-ViT-NAS
- Project page: OoD-ViT-NAS
- Downloads last month
- 19