Domain adaptation from multiple sources via auxiliary classifiers

By Lixin Duan, Ivor Wai-Hung Tsang, Dong Xu, Tat-seng Chua. 2009

338 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Representation & Self-Supervised Learning, Signal Processing

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Representation & Self-Supervised Learning

In Signal Processing

Citations in the corpus

Algorithms and Theory for Multiple-Source Adaptation (2018, background): “Duan et al. (2009, 2012) considered a somewhat similar problem where the learner leverages pre-trained predictors from the source domains to learn a good predictor on the target domain. However, they assume plenty of unlabeled target data to form a meaningful regularizer and they also assume a small”
Domain Generalization via Conditional Invariant Representations (2018, methodology): “Domain generalization has been widely applied in classi-ﬁcation tasks (Xu et al. 2014; Duan et al. 2009; Muan-det, Balduzzi, and Sch ¨ olkopf 2013; Ghifary et al. 2017; Ghifary et al. 2015; Erfani et al. 2016).”
Multi-Source Domain Adaptation: A Causal View (2015, background): “( Gao et al. 2008), (Duan et al. 2009), and (Chattopad-hyay et al. 2011) combine the predictions made by the source hypotheses, with the weights determined in different ways.”
A Comprehensive Survey on Transfer Learning (2019, methodology): “The objective function of each source classifier, denoted by fSk (with k = 1, . . . , mS), is similar to that of DAM, which is presented as follows: min fS k − nSk i=1 log P ySki |x Sk i ; f S k + λ2Χ fSk +λ1 nT,U i=1 yj∈Y S 1 mS mS k0=1 P yj |xT,Ui ; f S k0 where fSk denotes the decision function c”
Learning Multi-domain Convolutional Neural Networks for Visual Tracking (2015, methodology): “[8] introduced a domain-weighted combination of SVMs for video concept detection, and Hoffman et al.”
Domain Adaptation for Visual Applications: A Comprehensive Survey (2017, background): “Furthermore, in the case of multi-source domain adaptation [143, 67, 227, 265, 220, 116, 200, 102, 107, 52] different source specific tasks are jointly exploited in the interest of the target task.”
Return of Frustratingly Easy Domain Adaptation (2015, methodology): “In Table 1, we compare our method to five recent published methods: SVMA (Duan, Tsang, and Xu 2012), DAM (Duan et al. 2009), GFK (Gong et al.”
A Deeper Look at Dataset Bias (2015, methodology): “In Figure 4 we present the results obtained using DECAF7 with the landmark method, and the reshape approach combined respectively with SA and Domain Adaptation Machine (DAM [8]).”
More Classifiers, Less Forgetting: A Generic Multi-classifier Paradigm for Incremental Learning (2020, background): “On the one hand, the research objective is to maximize the consensus of outputs from multiple classifiers, to consolidate the transfer learning from source domain to target domain [27,8].”
Domain Adaptation on the Statistical Manifold (2014, background): “Several semi-supervised approaches have also been designed to handle the case where multiple source domains are available [11, 21].”
Understanding Deep Learning Techniques for Image Segmentation (2019, methodology): “Since then, it has been improved significantly with features like layer-wise training [13], rectified linear activations [153], batch normalization [84], auxiliary classifiers [52], atrous convolutions [211], skip connections [78], better optimization techniques [97] and so on.”
Structure-Preserved Multi-source Domain Adaptation (2016, background): “Although there are studies on multisource domain adaptation [4], most of them require target labels for classifier adaptation, which is different from our problem setting, i.e., unsupervised domain adaptation.”
Med3D: Transfer Learning for 3D Medical Image Analysis (2019, methodology): “Duan et al. [26] proposed a data-dependent regularizer based on support vector machine for video concept detection.”
Generalized Transfer Subspace Learning Through Low-Rank Constraint (2014, methodology): “More and more research effort is being devoted to visual domain adaptation (Yang et al. 2007; Jiang et al. 2008; Wang et al. 2008; Duan et al. 2009, 2012a,b,c; Si et al. 2010; Saenko et al. 2010; Gopalan et al. 2011; Kulis et al. 2011; Aytar and Zisserman 2011; Qi et al. 2011; Lim et al. 2011; Gong…”
Learning the Roots of Visual Domain Shift (2016, background): “Machine (DAM) originally introduced in [8].”
Heterogeneous Domain Generalization Via Domain Mixup (2020, background): “Domain adaptation (DA) [2, 3, 4, 5] assumes that we can obtain the samples from the testing data (a.”