FRIDA -- Generative Feature Replay for Incremental Domain Adaptation. (arXiv:2112.14316v2 [cs.CV] UPDATED)

We tackle the novel problem of incremental unsupervised domain adaptation
(IDA) in this paper. We assume that a labeled source domain and different
unlabeled target domains are incrementally observed with the constraint that
data corresponding to the current domain is only available at a time. The goal
is to preserve the accuracies for all the past domains while generalizing well
for the current domain. The IDA setup suffers due to the abrupt differences
among the domains and the unavailability of …

arxiv cv domain adaptation incremental