Continuous Domain Adaptation with Variational Domain-Agnostic Feature Replay

By Qicheng Lao, Xiang Jiang, Mohammad Havaei, Yoshua Bengio. 2020

34 citations recorded by the source corpus.

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.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Replay in Deep Learning: Current Approaches and Missing Biological Elements (2021, neutral): “…representational replay (Hayes et al., 2020; Caccia et al., 2020; Pellegrini et al., 2019; Iscen et al., 2020) or generative representational replay (van de Ven et al., 2020; Lao et al., 2020), instead of veridical (raw pixel) replay, have been proposed to improve continual learning performance.”
Continual Test-Time Domain Adaptation (2022, background): “Ideas from continual learning are adopted for continuous domain adaptation approaches [3, 30].”
Learning to Adapt to Evolving Domains (2020, background): “[3, 15] learn discrete target tasks online, but they do not address the evolving nature of target domains to harness it in knowledge transfer.”
Gradient Regularized Contrastive Learning for Continual Domain Adaptation (2020, background): “When learning a sequence of unlabeled target domains, continual domain adaptation aims to achieve good generalization abilities on all seen domains (Lao et al. 2020).”