Continuous Domain Adaptation with Variational Domain-Agnostic Feature Replay

TL;DR

This paper introduces a variational domain-agnostic feature replay method for continuous domain adaptation, enabling models to learn new tasks in changing environments while retaining prior knowledge.

Contribution

It proposes a novel approach combining inference, generative, and solver modules to handle domain and task drifts in non-stationary environments.

Findings

Effective in handling both domain and task drifts

Maintains performance on previous tasks during adaptation

Applicable to practical continuous learning scenarios

Abstract

Learning in non-stationary environments is one of the biggest challenges in machine learning. Non-stationarity can be caused by either task drift, i.e., the drift in the conditional distribution of labels given the input data, or the domain drift, i.e., the drift in the marginal distribution of the input data. This paper aims to tackle this challenge in the context of continuous domain adaptation, where the model is required to learn new tasks adapted to new domains in a non-stationary environment while maintaining previously learned knowledge. To deal with both drifts, we propose variational domain-agnostic feature replay, an approach that is composed of three components: an inference module that filters the input data into domain-agnostic representations, a generative module that facilitates knowledge transfer, and a solver module that applies the filtered and transferable knowledge to solve the queries. We address the two fundamental scenarios in continuous domain adaptation, demonstrating the effectiveness of our proposed approach for practical usage.