{"url":"/method/diffaugment","slug":"diffaugment","name":"DiffAugment","full_name":"DiffAugment","full_name_withheld":false,"description_markdown":"**Differentiable Augmentation (DiffAugment)** is a set of differentiable image transformations used to augment data during [GAN](https://paperswithcode.com/method/gan) training. The transformations are applied to the real and generated images. It enables the gradients to be propagated through the augmentation back to the generator, regularizes\r\nthe discriminator without manipulating the target distribution, and maintains the balance of training\r\ndynamics. Three choices of transformation are preferred by the authors in their experiments: Translation, [CutOut](https://paperswithcode.com/method/cutout), and Color.","description_state":"present","introduced_year":null,"introduced_by":{"title":"Differentiable Augmentation for Data-Efficient GAN Training","paper":"/paper/differentiable-augmentation-for-data","first_author":"Shengyu Zhao","n_authors":5,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/differentiable-augmentation-for-data"},"source":{"url":"https://arxiv.org/abs/2006.10738v4","title":"Differentiable Augmentation for Data-Efficient GAN Training","url_on_a_paper_host":true},"code_snippet_url":"https://github.com/mit-han-lab/data-efficient-gans/blob/ed7e725ae83c7bb4d2b0eace558ba1609d098e66/DiffAugment_pytorch.py#L9","code_snippet_url_on_a_code_host":true,"categories":[{"area":"Computer Vision","area_id":"computer-vision","collection":"Adversarial Image Data Augmentation","url":"/methods/category/adversarial-image-data-augmentation","pwc_aliases":[]},{"area":"General","area_id":"general","collection":"Adversarial Training","url":"/methods/category/adversarial-training","pwc_aliases":[]}],"n_papers_tagged":3,"archive_num_papers":3,"papers_newest_first":[{"paper":null,"title":"DiffAugment: Diffusion based Long-Tailed Visual Relationship Recognition","date":"2024-01-01","arxiv_id":"2401.01387","n_code_links":0,"syntology":null},{"paper":"/paper/importance-of-feature-extraction-in-the","title":"Feature Extraction for Generative Medical Imaging Evaluation: New Evidence Against an Evolving Trend","date":"2023-11-22","arxiv_id":"2311.13717","n_code_links":2,"syntology":null},{"paper":"/paper/differentiable-augmentation-for-data","title":"Differentiable Augmentation for Data-Efficient GAN Training","date":"2020-06-18","arxiv_id":"2006.10738","n_code_links":13,"syntology":{"ran":36,"of":58,"unverified":22,"pointer_only":12}}],"papers_shown":3,"tasks":[{"task":"/task/data-augmentation","name":"Data Augmentation","papers":1},{"task":"/task/image-generation","name":"Image Generation","papers":1},{"task":"/task/medical-image-generation","name":"Medical Image Generation","papers":1},{"task":"/task/object","name":"Object","papers":1},{"task":null,"name":"Relation","papers":1},{"task":null,"name":"Triplet","papers":1}],"tasks_shown":6,"n_tasks":6,"usage_by_year":[{"year":"2020","papers":1},{"year":"2023","papers":1},{"year":"2024","papers":1}],"row_source":"methods_table","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/diffaugment"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}