Filters








123,942 Hits in 6.0 sec

Transfer Neural Trees for Heterogeneous Domain Adaptation [chapter]

Wei-Yu Chen, Tzu-Ming Harry Hsu, Yao-Hung Hubert Tsai, Yu-Chiang Frank Wang, Ming-Syan Chen
2016 Lecture Notes in Computer Science  
Heterogeneous domain adaptation (HDA) addresses the task of associating data not only across dissimilar domains but also described by different types of features.  ...  Moreover, to address semi-supervised HDA, a unique embedding loss term for preserving prediction and structural consistency between targetdomain data is introduced into TNT.  ...  With the embedding loss for enforcing prediction and structural consistency between target-domain data, plus the use of our Transfer-NDF with stochastic pruning for adapting representative neurons, our  ... 
doi:10.1007/978-3-319-46454-1_25 fatcat:c47z35dy4nbt3l76cdqqmqmtei

DA-HGT: Domain Adaptive Heterogeneous Graph Transformer [article]

Tiancheng Huang, Ke Xu, Donglin Wang
2021 arXiv   pre-print
In this paper, we investigate Heterogeneous Information Networks (HINs) with partially shared node types and propose a novel Domain Adaptive Heterogeneous Graph Transformer (DA-HGT) to handle the domain  ...  Domain adaptation using graph networks learns label-discriminative and network-invariant node embeddings by sharing graph parameters.  ...  It ensures the consistency of label space between the source and target domains.  ... 
arXiv:2012.05688v2 fatcat:tsqr3nj47bhuze3maemr25ncfe

Unsupervised Domain Adaptation with Semantic Consistency across Heterogeneous Modalities for MRI Prostate Lesion Segmentation [article]

Eleni Chiou, Francesco Giganti, Shonit Punwani, Iasonas Kokkinos, Eleftheria Panagiotaki
2021 arXiv   pre-print
In our work we rely on stochastic generative modeling to translate across two heterogeneous domains at pixel space and introduce two new loss functions that promote semantic consistency.  ...  When compared to several unsupervised domain adaptation approaches, our approach yields substantial improvements, that consistently carry over to the semi-supervised and supervised learning settings.  ...  Introduction Domain adaptation transfers knowledge from a label-rich 'source' domain to a label-scarce or unlabeled 'target' domain.  ... 
arXiv:2109.09736v1 fatcat:b5gajmildjb6vddyfep3fskkcy

Locality Preserving Joint Transfer for Domain Adaptation [article]

Li Jingjing and Jing Mengmeng and Lu Ke and Zhu Lei and Shen Heng Tao
2019 arXiv   pre-print
Domain adaptation aims to leverage knowledge from a well-labeled source domain to a poorly-labeled target domain.  ...  Notably, our approach is suitable for both homogeneous and heterogeneous domain adaptation by learning domain-specific projections.  ...  As an extension of homogeneous domain adaptation, heterogeneous methods can handle domains with arbitrary features and dimensionalities.  ... 
arXiv:1906.07441v1 fatcat:2ontem74c5cvlo2qditieesx2u

Heterogeneous Domain Adaptation via Soft Transfer Network [article]

Yuan Yao, Yu Zhang, Xutao Li, Yunming Ye
2019 arXiv   pre-print
Heterogeneous domain adaptation (HDA) aims to facilitate the learning task in a target domain by borrowing knowledge from a heterogeneous source domain.  ...  Further, STN introduces an adaptive coefficient to gradually increase the importance of the soft-labels since they will become more and more accurate as the number of iterations increases.  ...  [36] propose a Label and Structure-consistent Unilateral Projection (LS-UP) model.  ... 
arXiv:1908.10552v1 fatcat:owcelg242fhb5mbxxbjigijwpe

Cross-Domain Structure Preserving Projection for Heterogeneous Domain Adaptation [article]

Qian Wang, Toby P. Breckon
2021 arXiv   pre-print
Heterogeneous Domain Adaptation (HDA) addresses the transfer learning problems where data from the source and target domains are of different modalities (e.g., texts and images) or feature dimensions (  ...  Traditional domain adaptation algorithms assume that the representations of source and target samples reside in the same feature space, hence are likely to fail in solving the heterogeneous domain adaptation  ...  Figure 1 : 1 An illustration of the heterogeneous domain adaptation problem and our proposed approach using cross-domain structure preserving projection.  ... 
arXiv:2004.12427v3 fatcat:rx5dgopfjbhkxpf6eua3itxywy

Multi-source Heterogeneous Domain Adaptation with Conditional Weighting Adversarial Network [article]

Yuan Yao, Xutao Li, Yu Zhang, Yunming Ye
2021 arXiv   pre-print
Heterogeneous domain adaptation (HDA) tackles the learning of cross-domain samples with both different probability distributions and feature representations.  ...  The proposed CWAN adversarially learns a feature transformer, a label classifier, and a domain discriminator.  ...  Note that the label space of N is consistent with that of E. Fig. 8 shows the weights of different source domains on the tasks of E, F, N → S and G, I, N → S in the last iteration.  ... 
arXiv:2008.02714v2 fatcat:oc7zcidyo5gp5okghohqjuno64

Simultaneous Semantic Alignment Network for Heterogeneous Domain Adaptation [article]

Shuang Li, Binhui Xie, Jiashu Wu, Ying Zhao, Chi Harold Liu, Zhengming Ding
2020 arXiv   pre-print
Heterogeneous domain adaptation (HDA) transfers knowledge across source and target domains that present heterogeneities e.g., distinct domain distributions and difference in feature type or dimension.  ...  Notably, a pseudo-label refinement procedure with geometric similarity involved is introduced to enhance the target pseudo-label assignment accuracy.  ...  THE PROPOSED ALGORITHM 3.1 Preliminary and Motivation For heterogeneous domain adaptation (HDA) with a semi-supervised setting, we have one labeled source domain and one scarcely labeled target domain.  ... 
arXiv:2008.01677v2 fatcat:my3hq5hbmnhqjkeoyo7p3dlh24

BoMuDANet: Unsupervised Adaptation for Visual Scene Understanding in Unstructured Driving Environments [article]

Divya Kothandaraman, Rohan Chandra, Dinesh Manocha
2021 arXiv   pre-print
Our method is designed for unstructured real-world scenarios with dense and heterogeneous traffic consisting of cars, trucks, two-and three-wheelers, and pedestrians.  ...  We describe a new semantic segmentation technique based on unsupervised domain adaptation (DA), that can identify the class or category of each region in RGB images or videos.  ...  Architecture: Consistent with adversarial domain adaptation [42] , our network consists of a DNN for semantic segmentation, and domain discriminators.  ... 
arXiv:2010.03523v3 fatcat:72sr7avhhjb7niwkheizlc2eeu

Asymmetric Heterogeneous Transfer Learning: A Survey

Magda Friedjungová, Marcel Jiřina
2017 Proceedings of the 6th International Conference on Data Science, Technology and Applications  
This overview focuses on the current progress in the new and unique area of transfer learning -asymmetric heterogeneous transfer learning.  ...  One of the main prerequisites in most machine learning and data mining tasks is that all available data originates from the same domain.  ...  Transductive learning also assumes that we have labeled source data and unlabeled target data, but heterogeneous transfer learning is able to work with different combinations of labeled and unlabeled data  ... 
doi:10.5220/0006396700170027 dblp:conf/data/FriedjungovaJ17 fatcat:xxogibknanf6rkj65cv5en7v5e

Distance Metric Facilitated Transportation between Heterogeneous Domains

Han-Jia Ye, Xiang-Rong Sheng, De-Chuan Zhan, Peng He
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
In this paper, we focus on transferring between heterogeneous domains, i.e., those with different feature spaces, and propose the Metric Transporation on HEterogeneous REpresentations (MapHere) approach  ...  Most existing methods restrict tasks connection on the same feature sets, or require aligned examples cross domains, even cannot take full advantage of the limited label information.  ...  Emphasizing domain linkages, MAPHERE transfers label and structure information across domain effectively.  ... 
doi:10.24963/ijcai.2018/418 dblp:conf/ijcai/YeSZH18 fatcat:ajokrlayerf35ct2mixtmo23le

Semi-Supervised Optimal Transport for Heterogeneous Domain Adaptation

Yuguang Yan, Wen Li, Hanrui Wu, Huaqing Min, Mingkui Tan, Qingyao Wu
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
To match the samples between heterogeneous domains, we propose to preserve the semantic consistency between heterogeneous domains by incorporating label information into the entropic Gromov-Wasserstein  ...  Heterogeneous domain adaptation (HDA) aims to exploit knowledge from a heterogeneous source domain to improve the learning performance in a target domain.  ...  Acknowledgments This work was supported by National Natural Science Foundation of China (NSFC) 61502177 and 61602185, and Recruitment Program for Young Professionals, and Guangdong Provincial Scientific  ... 
doi:10.24963/ijcai.2018/412 dblp:conf/ijcai/Yan0WMTW18 fatcat:4iejfpaqmrbbdduidxzmcqpj34

A survey on heterogeneous transfer learning

Oscar Day, Taghi M. Khoshgoftaar
2017 Journal of Big Data  
Heterogeneous transfer learning is characterized by the source and target domains having differing feature spaces, but may also be combined with other issues such as differing data distributions and label  ...  Currently, most transfer learning methods assume the source and target domains consist of the same feature spaces which greatly limits their applications.  ...  Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.  ... 
doi:10.1186/s40537-017-0089-0 fatcat:bpfjycwlkrawzdyyfv2ugle5cy

Representation Learning with Multiple Lipschitz-Constrained Alignments on Partially-Labeled Cross-Domain Data

Songlei Jian, Liang Hu, Longbing Cao, Kai Lu
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
MULAN utilizes both unlabeled and labeled data in the source and target domains to address distribution heterogeneity by Lipschitz-constrained adversarial distribution alignment and structure heterogeneity  ...  MULAN shows its superior performance on partially-labeled semi-supervised domain adaptation and few-shot domain adaptation and outperforms the state-of-the-art visual domain adaptation models by up to  ...  domains and the data distributions and structures are usually highly heterogeneous in different domains.  ... 
doi:10.1609/aaai.v34i04.5856 fatcat:an2fveyfe5hl3f5fzvtgskrx7y

Deep Visual Domain Adaptation: A Survey [article]

Mei Wang, Weihong Deng
2018 arXiv   pre-print
Deep domain adaption has emerged as a new learning technique to address the lack of massive amounts of labeled data.  ...  In this paper, we provide a comprehensive survey of deep domain adaptation methods for computer vision applications with four major contributions.  ...  ONE-STEP DOMAIN ADAPTATION As mentioned in Section II-A, the data in the target domain have three types regardless of homogeneous or heterogeneous DA: 1) supervised DA with labeled data, 2) semi-supervised  ... 
arXiv:1802.03601v4 fatcat:d5hwwecipjfjzmh7725lmepzfe
« Previous Showing results 1 — 15 out of 123,942 results