A Generic Self-Supervised Framework of Learning Invariant Discriminative Features

TL;DR

This paper introduces a versatile self-supervised learning framework that uses a self-transformation mechanism and contrastive learning to generate invariant representations across diverse data types, outperforming existing autoencoder-based methods.

Contribution

It proposes a generic SSL framework with a self-transformation process that replaces prior data-specific transformations, enabling invariant feature learning across multiple data modalities.

Findings

Outperforms many state-of-the-art methods on various data types

Demonstrates robustness and pattern recognition across visual, audio, text, and mass spectrometry data

Effective in learning invariant representations without data-specific customization

Abstract

Self-supervised learning (SSL) has become a popular method for generating invariant representations without the need for human annotations. Nonetheless, the desired invariant representation is achieved by utilising prior online transformation functions on the input data. As a result, each SSL framework is customised for a particular data type, e.g., visual data, and further modifications are required if it is used for other dataset types. On the other hand, autoencoder (AE), which is a generic and widely applicable framework, mainly focuses on dimension reduction and is not suited for learning invariant representation. This paper proposes a generic SSL framework based on a constrained self-labelling assignment process that prevents degenerate solutions. Specifically, the prior transformation functions are replaced with a self-transformation mechanism, derived through an unsupervised training process of adversarial training, for imposing invariant representations. Via the self-transformation mechanism, pairs of augmented instances can be generated from the same input data. Finally, a training objective based on contrastive learning is designed by leveraging both the self-labelling assignment and the self-transformation mechanism. Despite the fact that the self-transformation process is very generic, the proposed training strategy outperforms a majority of state-of-the-art representation learning methods based on AE structures. To validate the performance of our method, we conduct experiments on four types of data, namely visual, audio, text, and mass spectrometry data, and compare them in terms of four quantitative metrics. Our comparison results indicate that the proposed method demonstrate robustness and successfully identify patterns within the datasets.