Differentiable Data Augmentation with Kornia
This work provides a review and benchmark for researchers and developers using differentiable data augmentation in PyTorch.
This paper reviews Kornia's differentiable data augmentation (DDA) module for 2D and 3D tensors, which integrates DA pipelines with PyTorch components. It also includes a benchmark of different DA frameworks and a review of approaches using Kornia DDA.
In this paper we present a review of the Kornia differentiable data augmentation (DDA) module for both for spatial (2D) and volumetric (3D) tensors. This module leverages differentiable computer vision solutions from Kornia, with an aim of integrating data augmentation (DA) pipelines and strategies to existing PyTorch components (e.g. autograd for differentiability, optim for optimization). In addition, we provide a benchmark comparing different DA frameworks and a short review for a number of approaches that make use of Kornia DDA.