Self-Guided Masked Autoencoders for Domain-Agnostic Self-Supervised Learning
This addresses the challenge of domain-specific assumptions in self-supervised learning for researchers in fields like biology and physics, though it is incremental as it builds on masked modeling.
The paper tackled the problem of extending self-supervised learning to new data modalities by proposing a fully domain-agnostic masked modeling method, achieving state-of-the-art performance on benchmarks in protein biology, chemical property prediction, and particle physics.
Self-supervised learning excels in learning representations from large amounts of unlabeled data, demonstrating success across multiple data modalities. Yet, extending self-supervised learning to new modalities is non-trivial because the specifics of existing methods are tailored to each domain, such as domain-specific augmentations which reflect the invariances in the target task. While masked modeling is promising as a domain-agnostic framework for self-supervised learning because it does not rely on input augmentations, its mask sampling procedure remains domain-specific. We present Self-guided Masked Autoencoders (SMA), a fully domain-agnostic masked modeling method. SMA trains an attention based model using a masked modeling objective, by learning masks to sample without any domain-specific assumptions. We evaluate SMA on three self-supervised learning benchmarks in protein biology, chemical property prediction, and particle physics. We find SMA is capable of learning representations without domain-specific knowledge and achieves state-of-the-art performance on these three benchmarks.