Learning to Transfer Examples for Partial Domain Adaptation

By Zhangjie Cao, Kaichao You, Mingsheng Long, Jianmin Wang, Qiang Yang. 2019

304 citations recorded by the source corpus.

Domain adaptation is critical for learning in new and unseen environments. With domain adversarial training, deep networks can learn disentangled and transferable features that effectively diminish the dataset shift between the source and target domains for knowledge transfer. In the era of Big Data, large-scale labeled datasets are readily available, stimulating the interest in partial domain adaptation (PDA), which transfers a recognizer from a large labeled domain to a small unlabeled domain. It extends standard domain adaptation to the scenario where target labels are only a subset of source labels. Under the condition that target labels are unknown, the key challenges of PDA are how to transfer relevant examples in the shared classes to promote positive transfer and how to ignore irrelevant ones in the source domain to mitigate negative transfer. In this work, we propose a unified approach to PDA, Example Transfer Network (ETN), which jointly learns domain-invariant representations across domains and a progressive weighting scheme to quantify the transferability of source examples. A thorough evaluation on several benchmark datasets shows that ETN consistently achieves state-of-the-art results for various partial domain adaptation tasks.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

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 Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Overcoming Negative Transfer: A Survey (2020, background): “Then Cao et al. [45] proposed example transfer network (ETN), which down-weights the irrelevant examples of outlier classes further on the source classiﬁer by adding the class-level weight proposed in [44] into classiﬁcation loss function.”
A Survey on Negative Transfer (2020, background): “If the source and target domains have different label spaces, then a TL algorithm that transfers the outlier source classes to the target domain would mis-align the features of outlier source classes to certain target classes, migrating harmful knowledge from the source domain to the target domain a”
Heterogeneous Domain Generalization Via Domain Mixup (2020, background): “Another related research direction is partial domain adaptation (PDA) [6], which assumes that the target label space is a subset of the source label space.”
Domain Consensus Clustering for Universal Domain Adaptation (2021, background): “Recently, ETN [5] proposes a progressive weighting scheme to estimate the transferability of source samples, while BA 3 US [ 36] incorporates an augmentation scheme and a complement entropy to avoid negative transfer and uncertainty propagation, respectively.”
Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation (2020, methodology): “…al., 2016; Ghifary et al., 2016) impose data reconstruction as an auxiliary task to ensure feature invariance, and some studies consider batch normalization (Cariucci et al., 2017; Wang et al., 2019) and adversarial dropout (Saito et al., 2018a; Lee et al., 2019b) within the network architecture.”
Universal Domain Adaptation through Self Supervision (2020, methodology): “ENT [3] is the state-of-the-art method in PDA.”
Unsupervised Domain Adaptation of Object Detectors: A Survey (2021, background): “Hence, it is important to mitigate the problem of negative transfer by down-weighting the data samples from classes that are not common between the domain, thereby promoting feature alignment only in the shared label spaces (positive transfer) [185], [186], [187], [188], [189].”
Source Data-Absent Unsupervised Domain Adaptation Through Hypothesis Transfer and Labeling Transfer (2020, methodology): “For partial-set DA tasks, we compare ours with IWAN [105], SAN [91], ETN [106], DRCN [107], RTNetadv [108], BA(3)US [109], and TSCDA [110].”
A Balanced and Uncertainty-aware Approach for Partial Domain Adaptation (2020, methodology): “41 ETN [6] 59.”
A Review of Single-Source Deep Unsupervised Visual Domain Adaptation (2020, background): “homogeneity homogeneous DA dS = dT source and target data are observed in the same space [11, 23, 36, 44–106, 113–151] heterogeneous DA dS 6= dT source and target data are observed in different spaces [107–112]”
Minimum Class Confusion for Versatile Domain Adaptation (2019, methodology): “Representative methods include SAN [2], IWAN [51], PADA [3] and ETN [4], introducing different weighting mechanisms to select out outlier source classes in the process of domain feature alignment.”
Collaborative Unsupervised Domain Adaptation for Medical Image Diagnosis (2019, background): “One widely accepted prerequisite of deep learning is rich annotated data [7].”