TTA-ImageNet-C

Mirror of ImageNet-C (Hendrycks & Dietterich, ICLR 2019) with a revision pin for reproducible test-time adaptation evaluation.

• Upstream: Zenodo record 2235448
• License: CC BY 4.0 (matches upstream)
• Maintained as part of: TTA-Evaluation-Harness

Citation

@inproceedings{hendrycks2019benchmarking,
  title={Benchmarking Neural Network Robustness to Common Corruptions and Perturbations},
  author={Hendrycks, Dan and Dietterich, Thomas},
  booktitle={ICLR},
  year={2019}
}

Structure

• 15 configs: one per corruption type (gaussian_noise, shot_noise, ..., jpeg_compression).
• 5 splits per config: severity_1 through severity_5, 50 000 images each (1000 classes x 50).
• Labels are ClassLabel with 1000 WordNet-ID names in torchvision order (lexicographic on wnid; n01440764 = idx 0 = tench).

Usage

from datasets import load_dataset

ds = load_dataset("WNJXYK/TTA-ImageNet-C",
                  name="gaussian_noise",
                  split="severity_5",
                  revision="v1.0")

Provenance

This mirror was built by scripts/publish_imagenetc.py in the TTA-Evaluation-Harness repo. JPEG bytes are copied 1:1 from the upstream files - no re-encoding, pixel-for-pixel identical to Hendrycks's release.