Filters








19,333 Hits in 6.5 sec

Identity Preserving Generative Adversarial Network for Cross-Domain Person Re-identification [article]

Jialun Liu
2018 arXiv   pre-print
Most existing person re-identification (re-ID) models often fail to generalize well from the source domain where the models are trained to a new target domain without labels, because of the bias between  ...  The proposed method has two excellent properties: 1) only a single model is employed to translate the labeled images from the source domain to the target camera domains in an unsupervised manner; 2) The  ...  The better solution of domain adaptation is to reduce the bias between source domain and each subdomain (camera domain) in target domain.  ... 
arXiv:1811.11510v1 fatcat:evzd2p56jrbqhkyyrkyedareum

Unsupervised Domain Expansion from Multiple Sources [article]

Jing Zhang, Wanqing Li, Lu sheng, Chang Tang, Philip Ogunbona
2020 arXiv   pre-print
Specifically, this paper presents a method for unsupervised multi-source domain expansion (UMSDE) where only the pre-learned models of the source domains and unlabelled new domain data are available.  ...  Given an existing system learned from previous source domains, it is desirable to adapt the system to new domains without accessing and forgetting all the previous domains in some applications.  ...  By contrast, our source model weights are used for reducing the bias among different domains.  ... 
arXiv:2005.12544v1 fatcat:q5f5yt4rwrayvdm5mwouftj3pq

Identity Preserving Generative Adversarial Network for Cross-Domain Person Re-identification

Jialun Liu, Wenhui Li, Hongbin Pei, Ying Wang, Feng Qu, You Qu, Yuhao Chen
2019 IEEE Access  
The previous methods directly reduce the bias by image-to-image style translation between the source and the target domain in an unsupervised manner.  ...  INDEX TERMS Person re-identification, domain adaptation, style transfer, unsupervised learning.  ...  ACKNOWLEDGMENT Jialun Liu would like to thank Zhun Zhong and Xiaohang Li for helpful discussions and encouragement.  ... 
doi:10.1109/access.2019.2933910 fatcat:llqipk4og5fbrlbbsqpetzgiiq

Denoised Maximum Classifier Discrepancy for Source-Free Unsupervised Domain Adaptation

Tong Chu, Yahao Liu, Jinhong Deng, Wen Li, Lixin Duan
2022 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Source-Free Unsupervised Domain Adaptation(SFUDA) aims to adapt a pre-trained source model to an unlabeled target domain without access to the original labeled source domain samples.  ...  In particular, we first minimize the distribution mismatch between the selected pseudo-labeled samples and the remaining target domain samples to alleviate the sample selection bias.  ...  Acknowledgements This work is supported by the Major Project for New Generation of AI under Grant No. 2018AAA0100400, the National Natural Science Foundation of China (Grant No. 62176047), Beijing Natural  ... 
doi:10.1609/aaai.v36i1.19925 fatcat:7xbbni5tijaovie623jo6j77bm

Imitating Targets from all sides: An Unsupervised Transfer Learning method for Person Re-identification [article]

Jiajie Tian, Zhu Teng, Rui Li, Yan Li, Baopeng Zhang, Jianping Fan
2021 arXiv   pre-print
In terms of this issue, given a labelled source training set and an unlabelled target training set, we propose an unsupervised transfer learning method characterized by 1) bridging inter-dataset bias and  ...  learn a discriminative representation across domains; 3) exploiting the underlying commonality across different domains from the class-style space to improve the generalization ability of re-ID models  ...  One solution to this problem is unsupervised domain adaption (UDA) where models are trained on a source domain consisting of labelled images and adapted on the target domain composed of unlabelled images  ... 
arXiv:1904.05020v2 fatcat:d7tmspdg4fbjngwdrhsaq5gy7a

Unsupervised Deep Domain Adaptation for Pedestrian Detection [article]

Lihang Liu, Weiyao Lin, Lisheng Wu, Yong Yu, Michael Ying Yang
2018 arXiv   pre-print
Meanwhile, we also reuse negative samples from the source domain to compensate for the imbalance between the amount of positive samples and negative samples.  ...  First, we utilize an iterative algorithm to iteratively select and auto-annotate positive pedestrian samples with high confidence as the training samples for the target domain.  ...  Domain adaptation aims to reduce the amount of data needed for the target domain. Many domain adaptation works try to learn a common representation space shared between the source and target domain.  ... 
arXiv:1802.03269v1 fatcat:wosnlbjegrbyfjbfhr24ta5nqu

Progressive Feature Alignment for Unsupervised Domain Adaptation [article]

Chaoqi Chen, Weiping Xie, Wenbing Huang, Yu Rong, Xinghao Ding, Yue Huang, Tingyang Xu, Junzhou Huang
2019 arXiv   pre-print
Unsupervised domain adaptation (UDA) transfers knowledge from a label-rich source domain to a fully-unlabeled target domain.  ...  Moreover, upon observing that a good domain adaptation usually requires a non-saturated source classifier, we consider a simple yet efficient way to retard the convergence speed of the source classification  ...  samples and the target samples become closer to The EHTS is biased to favor easier samples and this bias each other in the hidden space as training proceeds.  ... 
arXiv:1811.08585v2 fatcat:ezgoausuffcvxjoa7libj6k2oi

Adaptive Ensembling: Unsupervised Domain Adaptation for Political Document Analysis

Shrey Desai, Barea Sinno, Alex Rosenfeld, Junyi Jessy Li
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
To bridge this gap, we present adaptive ensembling, an unsupervised domain adaptation framework, equipped with a novel text classification model and time-aware training to ensure our methods work well  ...  In contrast, we can find corpora that label relevant documents but have limitations (e.g., from a single source or era), preventing their use for political science research.  ...  Thanks as well to Greg Durrett, Katrin Erk, and the anonymous reviewers for their helpful comments. This work was partially supported by the NSF Grant IIS-1850153.  ... 
doi:10.18653/v1/d19-1478 dblp:conf/emnlp/DesaiSRL19 fatcat:hhqzsequhbfezecnu62frzinsu

Heterogeneous Feature Space Based Task Selection Machine for Unsupervised Transfer Learning

Shan Xue, Jie Lu, Guangquan Zhang, Li Xiong
2015 2015 10th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)  
Since transfer learning cares more about relatedness between tasks and their domains, it is useful for handling massive data, which are not labeled, to overcome distribution and feature space gaps, respectively  ...  In this paper, we propose a new task selection algorithm in an unsupervised transfer learning domain, called as Task Selection Machine (TSM).  ...  However, it's not easy to tell target task from source tasks if we expect reducing computational complexity and when data are all unlabeled, i.e., to achieve unsupervised learning.  ... 
doi:10.1109/iske.2015.29 dblp:conf/iske/XueLZX15 fatcat:ondu6bwijfhdlkk6jkjwoibzpa

Adaptive Ensembling: Unsupervised Domain Adaptation for Political Document Analysis [article]

Shrey Desai, Barea Sinno, Alex Rosenfeld, Junyi Jessy Li
2019 arXiv   pre-print
To bridge this gap, we present adaptive ensembling, an unsupervised domain adaptation framework, equipped with a novel text classification model and time-aware training to ensure our methods work well  ...  In contrast, we can find corpora that label relevant documents but have limitations (e.g., from a single source or era), preventing their use for political science research.  ...  Thanks as well to Greg Durrett, Katrin Erk, and the anonymous reviewers for their helpful comments. This work was partially supported by the NSF Grant IIS-1850153.  ... 
arXiv:1910.12698v1 fatcat:2xdgjt3k6ndebiitkud7lqbshm

Unsupervised Vehicle Re-identification with Progressive Adaptation [article]

Jinjia Peng, Yang Wang, Huibing Wang, Zhao Zhang, Xianping Fu, Meng Wang
2020 arXiv   pre-print
For PAL, a data adaptation module is employed for source domain, which generates the images with similar data distribution to unlabeled target domain as "pseudo target samples".  ...  The existing methods heavily relied on well-labeled datasets for ideal performance, which inevitably causes fateful drop due to the severe domain bias between the training domain and the real-world scenes  ...  have similar distribution for the target domain, which reduces the bias between source and target domain. • Furthermore, the identity information of source domain is also preserved by turning the content  ... 
arXiv:2006.11486v1 fatcat:nlprfets5zbdrdn4qcvxg7cq5u

Unsupervised Vehicle Re-identification with Progressive Adaptation

Jinjia Peng, Yang Wang, Huibing Wang, Zhao Zhang, Xianping Fu, Meng Wang
2020 Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence  
For PAL, a data adaptation module is employed for source domain, which generates the images with similar data distribution to unlabeled target domain as "pseudo target samples".  ...  The existing methods heavily relied on well-labeled datasets for ideal performance, which inevitably causes fateful drop due to the severe domain bias between the training domain and the real-world scenes  ...  have similar distribution for the target domain, which reduces the bias between source and target domain. • Furthermore, the identity information of source domain is also preserved by turning the content  ... 
doi:10.24963/ijcai.2020/127 dblp:conf/ijcai/PengWWZFW20 fatcat:bwpc26sv2zbj3cf5ugmmc4c3cm

ACDC: Online Unsupervised Cross-Domain Adaptation [article]

Marcus de Carvalho, Mahardhika Pratama, Jie Zhang, Edward Yapp
2021 arXiv   pre-print
We propose ACDC, an adversarial unsupervised domain adaptation framework that handles multiple data streams with a complete self-evolving neural network structure that reacts to these defiances.  ...  We consider the problem of online unsupervised cross-domain adaptation, where two independent but related data streams with different feature spaces -- a fully labeled source stream and an unlabeled target  ...  layer depth, to say a few.  ... 
arXiv:2110.01326v1 fatcat:s2lb2jxpzjfhvobwscyliqfpgi

Content Preserving Image Translation with Texture Co-occurrence and Spatial Self-Similarity for Texture Debiasing and Domain Adaptation [article]

Myeongkyun Kang, Dongkyu Won, Miguel Luna, Philip Chikontwe, Kyung Soo Hong, June Hong Ahn, Sang Hyun Park
2022 arXiv   pre-print
Models trained on datasets with texture bias usually perform poorly on out-of-distribution samples since biased representations are embedded into the model.  ...  Recently, various image translation and debiasing methods have attempted to disentangle texture biased representations for downstream tasks, but accurately discarding biased features without altering other  ...  Unsupervised Domain Adaptation Unsupervised domain adaptation (UDA) methods in semantic segmentation have been proposed to address the domain shift problem where source and target domains are known in  ... 
arXiv:2110.07920v4 fatcat:2qmzsyfbkja3xdacigfdkkvmgq

Contradistinguisher: A Vapnik's Imperative to Unsupervised Domain Adaptation [article]

Sourabh Balgi, Ambedkar Dukkipati
2021 arXiv   pre-print
Recent domain adaptation works rely on an indirect way of first aligning the source and target domain distributions and then train a classifier on the labeled source domain to classify the target domain  ...  a direct approach to domain adaptation that does not require domain alignment.  ...  The authors would also like to thank anonymous reviewers for providing their valuable feedback that helped in improving the manuscript.  ... 
arXiv:2005.14007v3 fatcat:nkofjc3pbjdt5aq7arhdh3bmj4
« Previous Showing results 1 — 15 out of 19,333 results