LAMBO: Landmarks Augmentation With Manifold-Barycentric Oversampling

IEEE Access. 2022. No. 10. Article 3219934.

Bespalov Y., Buzun N., Kachan O., Dylov D.

We propose the first data augmentation method based on optimal transport theory, with the generated data being guaranteed to belong to the original data manifold. The proposed algorithm randomly samples a clique in the nearest-neighbors graph representing the data knowledge and computes the Wasserstein barycenter between the neighbours with random uniform weights. Being extremely natural- looking, many such barycenters are then produced iteratively to overpopulate the original dataset. We apply this approach to the problem of landmarks detection in unsupervised and semi-supervised scenarios in the popular tasks of face keypoints extraction, pose detection, and the segmentation of anatomical contours in medical imaging. The barycentric oversampling approach is shown to outperform state-of-the-art data augmentation methods. The code is available at https://github.com/cviaai/LAMBO/.

Keywords: manifold, Wasserstein barycenter, data augmentation, Generative Adversarial Networks (GANs), oversampling