Unified Deep Supervised Domain Adaptation and Generalization

By Saeid Motiian, Marco Piccirilli, D. Adjeroh, Gianfranco Doretto. 2017

889 citations recorded by the source corpus.

This work provides a unified framework for addressing the problem of visual supervised domain adaptation and generalization with deep models. The main idea is to exploit the Siamese architecture to learn an embedding subspace that is discriminative, and where mapped visual domains are semantically aligned and yet maximally separated. The supervised setting becomes attractive especially when only few target data samples need to be labeled. In this scenario, alignment and separation of semantic probability distributions is difficult because of the lack of data. We found that by reverting to point-wise surrogates of distribution distances and similarities provides an effective solution. In addition, the approach has a high “speed” of adaptation, which requires an extremely low number of labeled target training samples, even one per category can be effective. The approach is extended to domain generalization. For both applications the experiments show very promising results.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Representation & Self-Supervised Learning

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

In Representation & Self-Supervised Learning

Citations in the corpus

Learning Robust Global Representations by Penalizing Local Predictive Power (2019, methodology): “At a high level these methods comprise two subtypes: ﬁne-tuning over target domain (Long et al., 2016; Hoffman et al., 2017; Motiian et al., 2017a; Gebru et al., 2017; Volpi et al., 2018) and coercing domain invariance via adversarial learning (or further extensions) (Ganin et al., 2016; Bousmalis…”
Multiple-Source Adaptation with Domain Classifiers (2020, neutral): “…al., 2008, 2009a, Hoffman et al., 2018, Pan and Yang, 2010, Muandet et al., 2013, Xu et al., 2014, Hoffman et al., 2012, Gong et al., 2013a,b, Zhang et al., 2015, Ganin et al., 2016, Tzeng et al., 2015, Motiian et al., 2017b,a, Wang et al., 2019, Konstantinov and Lampert, 2019, Liu et al., 2015].”
A Theory of Multiple-Source Adaptation with Limited Target Labeled Data (2020, background): “Motiian et al. (2017b), Motiian et al. (2017a) and Wang et al. (2019) proposed to train maximally separated features via adversarial learning.”
Learning Robust Representations by Projecting Superficial Statistics Out (2018, methodology): “We compared the performance of HEX/ADV with several methods tested on this data including CAE (Rifai et al., 2011), MTAE (Ghifary et al., 2015), CCSA (Motiian et al., 2017), DANN (Ganin et al., 2016), Fusion (Mancini et al., 2018), LabelGrad, and CrossGrad (Shankar et al., 2018).”
Few-Shot Adversarial Domain Adaptation (2017, methodology): “Moreover, we report the accuracies obtained by fine-tuning (FT) the base model on available target data and also the recent work presented in [38].”
In Search of Lost Domain Generalization (2020, background): “These are Empirical Risk Minimization (ERM, Vapnik (1998)), Group Distributionally Robust Optimization (GroupDRO, Sagawa et al. (2019)), Inter-domain Mixup (Mixup, Xu et al.”
Generalizing to Unseen Domains via Adversarial Data Augmentation (2018, background): “Domain generalization methods [28, 22, 27, 33, 24] that propose different ways to better generalize to unknown domains are also related to our work.”
MSeg: A Composite Dataset for Multi-Domain Semantic Segmentation (2020, methodology): “We use a state-of-the-art multi-task learning algorithm [31] and a Domain Generalization (DG) algorithm [23].”
Domain Generalization Using a Mixture of Multiple Latent Domains (2019, methodology): “CCSA (Motiian et al. 2017): The deep model in mixture with the classification and contrastive semantic alignment loss to address supervised domain adaptation and generalization.”
Invariance-inducing regularization using worst-case transformations suffices to boost accuracy and spatial robustness (2019, methodology): “These approaches are based on augmenting the training data either statically [17, 30, 7, 45], ie.”
Single-Side Domain Generalization for Face Anti-Spoofing (2020, background): “[27] propose a new loss to guide the features of the same class to be close in the latent feature space.”
Adapting Object Detectors via Selective Cross-Domain Alignment (2019, neutral): “incorporating a domain classifier with gradient reversal [9], or directly reverting the distribution distances [33].”
Image-Image Domain Adaptation with Preserved Self-Similarity and Domain-Dissimilarity for Person Re-identification (2017, methodology): “Other methods seek to find a domain-invariant feature space [34, 31, 10, 30, 42, 11, 2].”
Best Sources Forward: Domain Generalization through Source-Specific Nets (2018, methodology): “This idea is exploited by previous methods searching for a domain invariant feature space where to project the data [5, 16] or by approaches attempting to generate domain agnostic classifiers using both shallow [18, 6] or deep learning models [10, 15, 12].”
FedDG: Federated Domain Generalization on Medical Image Segmentation via Episodic Learning in Continuous Frequency Space (2021, methodology): “[37] utilize a contrastive loss to minimize the distance between samples from the same class but different domains.”
Zero-shot Domain Adaptation without Domain Semantic Descriptors (2018, methodology): “CCSA is widely used zero-shot domain adaptation method that finds domain-invariant representations.”
Learning to Learn Single Domain Generalization (2020, methodology): “To reduce the discrepancy across domains, a series of methods are proposed for unsupervised [33, 43, 7, 38, 39] or supervised domain adaptation [31, 57].”