A Two-Stage Weighting Framework for Multi-Source Domain Adaptation

By Qian Sun, Rita Chattopadhyay, Sethuraman Panchanathan, Jieping Ye. 2011

221 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.

Citations in the corpus

Learngene: From Open-World to Your Learning Task (2021, background): “Transfer learning: Methods such as (Huang et al. 2006; Sugiyama et al. 2008; Sun et al. 2011; Shen et al. 2018; Lee et al. 2019; Duan et al. 2009; Zhuang et al. 2009; Evgeniou, Micchelli, and Pontil 2005), which focus on transferring the knowledge across domains, aim at improving the performance of ”
A Comprehensive Survey on Transfer Learning (2019, background): “proposed a multi-source framework termed 2-Stage Weighting Framework for Multi-Source Domain Adaptation (2SW-MDA) with the following two stages [29].”
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.”
Reshaping Visual Datasets for Domain Adaptation (2013, methodology): “Instead, we evaluate the effectiveness of our approach by investigating whether its automatically identified domains improve adaptation, that is, whether recognition accuracy on the target domains can be improved by reshaping the datasets into their latent source domains.”
Generalized Transfer Subspace Learning Through Low-Rank Constraint (2014, background): “…(Arnold et al. 2007; Dai et al. 2007a,b), and specifically, domain adaptation (Daumé III and Marcu 2006; Blitzer et al. 2006, 2011; Daumé 2007; Sun et al. 2011; Glorot et al. 2011; Chen et al. 2011; Lopez-Paz et al. 2012; Zhang et al. 2012) transforms the source domain by either manipulating…”
I Have Seen Enough: Transferring Parts Across Categories (2016, background): “Only a few papers consider DA from more than one source domain [47], and most from at most a few, while we consider 50.”
Joint cross-domain classification and subspace learning for unsupervised adaptation (2014, background): “Several approaches have been proposed for classification tasks in the unsupervised scenario, where no labeled target data are available.”
Transfer Feature Learning with Joint Distribution Adaptation (2013, methodology): “To reduce the difference between marginal distributions Ps(xs) and Pt(xt), we follow [8, 15, 23] and adopt the empirical Maximum Mean Discrepancy (MMD) as the distance measure to compare different distributions, which computes the distance between the sample means of the source and target data in th”
A Review of Single-Source Deep Unsupervised Visual Domain Adaptation (2020, background): “#targets single-target DA NT = 1 there is only one target domain [11, 23, 36, 44–112, 115–151] multi-target DA NT > 1 there are multiple target domains [113, 114]”
Agnostic Domain Generalization (2018, background): “Several other works followed this direction presenting algorithms based on the combination of source hypotheses [42, 9].”
Extracting Relationships by Multi-Domain Matching (2018, background): “A key idea from prior work is to match every source domain’s feature-space distribution to that of the target domain [37, 29].”
On handling negative transfer and imbalanced distributions in multiple source transfer learning (2014, background): “developed [2–5,7,8] to solve the problem of MSTL.”
Flexible Transfer Learning under Support and Model Shift (2014, result): “Similar to work on covariate shift, [15] weighted the samples from the source domain to deal with domain adaptation.”
Towards more Reliable Transfer Learning (2018, methodology): “A more sophisticated twostage weighting methodology based on distribution matching and conditional probability differences is presented in [28].”
A survey of multi-source domain adaptation (2015, methodology): “[30] proposed a two-stage domain adaptation methodology for multi-source problems.”
Domain Aggregation Networks for Multi-Source Domain Adaptation (2019, background): “Sun et al. (2011) developed a generalization bound but resorted to heuristic algorithms to adjust distribution shifts.”
Ieee Transactions on Pattern Analysis and Machine Intelligence Classification with Noisy Labels by Importance Reweighting (unknown): “Although these methods introduce approximation errors for learning the weight β * (X, ˆ Y ), their efficiency has been widely and empirically proven [48]–[50].”