A Comprehensive Survey on Domain Adaptation for Visual Applications

By Gabriella Csurka. 2017

330 citations recorded by the source corpus.

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

Citations in the corpus

Conditional variance penalties and domain shift robustness (2017, background): “Issues can arise when the learned representations rely on dependencies that vanish in test distributions, see for example Quionero-Candela et al. (2009); Torralba and Efros (2011); Csurka (2017) and references therein.”
Class-incremental Learning via Deep Model Consolidation (2019, background): “Domain adaptation or inductive transfer learning [10] aims to transfer and reuse knowledge in labeled out-of-domain instances.”
Continual Learning for Robotics (2019, background): “In computer vision, transferring what has been learned from a past environment to new environments would be often referred to as domain adaptation [118, 31].”
Deep Domain Adaptation in Action Space (2018, background): “In a recent survey paper [5], domain adaptation and transfer learning techniques have been comprehensively discussed with a speciﬁc view on visual applications.”
Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation (2020, background): “HTL (Kuzborskij & Orabona, 2013) can be seen as a generalization of parameter adaptation (Csurka, 2017), which assumes the optimal target hypothesis to be closely related to the source hypothesis.”
Unsupervised Multi-Target Domain Adaptation: An Information Theoretic Approach (2018, methodology): “Existing DA methods usually tackle the adaptation problems in one of the two settings: (semi)supervised DA [9] and unsupervised DA [10].”
Source Data-Absent Unsupervised Domain Adaptation Through Hypothesis Transfer and Labeling Transfer (2020, methodology): “nition [2], [4], [5], [6], semantic segmentation [3], [7], [8], [9], sentiment classification [10], [11], object detection [12], [13], and person re-identification [14], [15].”
Domain Adaptation for Medical Image Analysis: A Survey (2021, neutral): “In terms of whether the learning model is shallow or deep, the DA methods can be divided into shallow DA and deep DA [9], [10], [13].”
FDA: Fourier Domain Adaptation for Semantic Segmentation (2020, background): “Domain Adaptation aims to reduce the shift between two distributions [32, 10, 46].”
Unsupervised Domain Adaptation with Similarity Learning (2017, background): “Unsupervised Domain Adaptation: Non-deep approaches for unsupervised domain adaptation usually consist of matching the feature distribution between the source and target domain [24, 11].”
A Survey of Unsupervised Deep Domain Adaptation (2018, background): “cal use of domain adaptation to remove the reliance on potentially costly target data labels, will be the focus of this survey. A number of surveys have been created on the topic of domain adaptation [12, 24, 42, 48, 49, 121, 122, 162, 182, 227, 246, 285] and more generally transfer learning [45, 12”
Guarding Against Adversarial Domain Shifts with Counterfactual Regularization (2017, background): “M L ] 3 1 O ct 2 01 Problems arise when the learned representations rely on dependencies that vanish in test distributions (e.g. see Csurka (2017) and references therein).”
Learning to Diversify for Single Domain Generalization (2021, background): “The difference between the distributions is termed as the domain shift [36, 8].”
DCAN: Dual Channel-wise Alignment Networks for Unsupervised Scene Adaptation (2018, background): “There is a large body of work on domain adaptation (see [25,26] for a survey), and here we focus only on the most relevant literatures.”
Temporal Attentive Alignment for Large-Scale Video Domain Adaptation (2019, methodology): “Domain adaptation (DA) [31] has been studied extensively in recent years [5] to address the domain shift problem [36, 33], which means the models trained on source labeled dataset do not generalize well to target datasets and tasks.”