Augmentation Invariant Manifold Learning

TL;DR

This paper introduces a new theoretical framework and method for invariant manifold learning that leverages data augmentation to improve representation quality and downstream classification, supported by theoretical guarantees and experiments.

Contribution

It develops a statistical framework on a low-dimensional manifold for augmentation invariant learning and proposes a novel, efficient algorithm with theoretical guarantees.

Findings

Data augmentation enhances downstream classification performance.

The proposed method exploits geometric and invariant properties of data.

More complex augmentations lead to greater improvements.

Abstract

Data augmentation is a widely used technique and an essential ingredient in the recent advance in self-supervised representation learning. By preserving the similarity between augmented data, the resulting data representation can improve various downstream analyses and achieve state-of-the-art performance in many applications. Despite the empirical effectiveness, most existing methods lack theoretical understanding under a general nonlinear setting. To fill this gap, we develop a statistical framework on a low-dimension product manifold to model the data augmentation transformation. Under this framework, we introduce a new representation learning method called augmentation invariant manifold learning and design a computationally efficient algorithm by reformulating it as a stochastic optimization problem. Compared with existing self-supervised methods, the new method simultaneously exploits the manifold's geometric structure and invariant property of augmented data and has an explicit theoretical guarantee. Our theoretical investigation characterizes the role of data augmentation in the proposed method and reveals why and how the data representation learned from augmented data can improve the $k$-nearest neighbor classifier in the downstream analysis, showing that a more complex data augmentation leads to more improvement in downstream analysis. Finally, numerical experiments on simulated and real data sets are presented to demonstrate the merit of the proposed method.