Filters








49,624 Hits in 5.8 sec

Universal-RCNN: Universal Object Detector via Transferable Graph R-CNN

Hang Xu, Linpu Fang, Xiaodan Liang, Wenxiong Kang, Zhenguo Li
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Finally, an Inter-Domain Transfer Module is proposed to exploit diverse transfer dependencies across all domains and enhance the regional feature representation by attending and transferring semantic contexts  ...  Then an Intra-Domain Reasoning Module learns and propagates the sparse graph representation within one dataset guided by a spatial-aware GCN.  ...  Learning the Graph from Features. This scheme is almost the same as the one used in the Intra-Domain Reasoning Module. The visual feature is transformed to the latent space z.  ... 
doi:10.1609/aaai.v34i07.6937 fatcat:vlulmyim2bgxjekytexsmviube

Universal-RCNN: Universal Object Detector via Transferable Graph R-CNN [article]

Hang Xu, Linpu Fang, Xiaodan Liang, Wenxiong Kang, Zhenguo Li
2020 arXiv   pre-print
Finally, an InterDomain Transfer Module is proposed to exploit diverse transfer dependencies across all domains and enhance the regional feature representation by attending and transferring semantic contexts  ...  Then an Intra-Domain Reasoning Module learns and propagates the sparse graph representation within one dataset guided by a spatial-aware GCN.  ...  Learning the Graph from Features. This scheme is almost the same as the one used in the Intra-Domain Reasoning Module. The visual feature is transformed to the latent space z.  ... 
arXiv:2002.07417v1 fatcat:jju32xvo5zbe3ma3bzpzz6r2ku

Graphonomy: Universal Image Parsing via Graph Reasoning and Transfer [article]

Liang Lin and Yiming Gao and Ke Gong and Meng Wang and Xiaodan Liang
2021 arXiv   pre-print
In particular, Graphonomy learns the global and structured semantic coherency in multiple domains via semantic-aware graph reasoning and transfer, enforcing the mutual benefits of the parsing across domains  ...  The former extracts the semantic graph in each domain to improve the feature representation learning by propagating information with the graph; the latter exploits the dependencies among the graphs from  ...  domains via graph transfer learning to achieve multiple levels of human parsing tasks.  ... 
arXiv:2101.10620v1 fatcat:hnbuqiugsfhvbc7phn5htmsvcy

Graphonomy: Universal Human Parsing via Graph Transfer Learning

Ke Gong, Yiming Gao, Xiaodan Liang, Xiaohui Shen, Meng Wang, Liang Lin
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
In particular, Graphonomy first learns and propagates compact high-level graph representation among the labels within one dataset via Intra-Graph Reasoning, and then transfers semantic information across  ...  This poses many fundamental learning challenges, e.g. discovering underlying semantic structures among different label granularity, performing proper transfer learning across different image domains, and  ...  Additionally, we build an Inter-Graph Transfer module to attentively distill related semantics from the graph in one domain/task to the one in another domain, which bridges the semantic labels from different  ... 
doi:10.1109/cvpr.2019.00763 dblp:conf/cvpr/Gong0LS0L19 fatcat:rv5xgqag4jcghmzffi3s67fp2u

Graphonomy: Universal Human Parsing via Graph Transfer Learning [article]

Ke Gong, Yiming Gao, Xiaodan Liang, Xiaohui Shen, Meng Wang, Liang Lin
2019 arXiv   pre-print
In particular, Graphonomy first learns and propagates compact high-level graph representation among the labels within one dataset via Intra-Graph Reasoning, and then transfers semantic information across  ...  This poses many fundamental learning challenges, e.g. discovering underlying semantic structures among different label granularity, performing proper transfer learning across different image domains, and  ...  Additionally, we build an Inter-Graph Transfer module to attentively distill related semantics from the graph in one domain/task to the one in another domain, which bridges the semantic labels from different  ... 
arXiv:1904.04536v1 fatcat:di2yce3ytbhadml5lljt7yn66m

Graph Distillation for Action Detection with Privileged Modalities [article]

Zelun Luo, Jun-Ting Hsieh, Lu Jiang, Juan Carlos Niebles, Li Fei-Fei
2018 arXiv   pre-print
Common methods in transfer learning do not take advantage of the extra modalities potentially available in the source domain.  ...  On the other hand, previous work on multimodal learning only focuses on a single domain or task and does not handle the modality discrepancy between training and testing.  ...  Fig. 4 shows example distillation graphs learned on NTU RGB+D. The results show that our method, without transfer learning, is effective for action classification in the source domain.  ... 
arXiv:1712.00108v2 fatcat:vbvqueff4faahg365uzzgoffoq

Learning Robust Data Representation: A Knowledge Flow Perspective [article]

Zhengming Ding and Ming Shao and Handong Zhao and Sheng Li
2020 arXiv   pre-print
It is always demanding to learn robust visual representation for various learning problems; however, this learning and maintenance process usually suffers from noise, incompleteness or knowledge domain  ...  Thus, robust representation learning by removing noisy features or samples, complementing incomplete data, and mitigating the distribution difference becomes the key.  ...  ., visual features and semantic features that are highly coupled.  ... 
arXiv:1909.13123v2 fatcat:wll23rkrznejvhzsihc6rwcwve

A Survey on Visual Transfer Learning using Knowledge Graphs [article]

Sebastian Monka, Lavdim Halilaj, Achim Rettinger
2022 arXiv   pre-print
This survey focuses on visual transfer learning approaches using KGs.  ...  We explain the notion of feature extractor, while specifically referring to visual and semantic features.  ...  Zero-Shot Learning is a visual transfer learning task with labeled source domain data and unlabeled target domain data.  ... 
arXiv:2201.11794v1 fatcat:tapql5h4j5dvrnxjkaxek2cquu

A survey on visual transfer learning using knowledge graphs

Sebastian Monka, Lavdim Halilaj, Achim Rettinger, Mehwish Alam, Davide Buscaldi, Michael Cochez, Francesco Osborne, Diego Reforgiato Recupero, Harald Sack
2022 Semantic Web Journal  
This survey focuses on visual transfer learning approaches using KGs, as we believe that KGs are well suited to store and represent any kind of auxiliary knowledge.  ...  We explain the notion of feature extractor, while specifically referring to visual and semantic features.  ...  Acknowledgements This publication was created as part of the research project "KI Delta Learning" (project number: 19A19013D) funded by the Federal Ministry for Economic Affairs and Energy (BMWi) on the  ... 
doi:10.3233/sw-212959 fatcat:f4s43if3nbcxxfvrbtpdrrs2ry

Semi-supervised Breast Histological Image Classification by Node-attention Graph Transfer Network

Liheng Gong, Jingjing Yang, Xiao Zhang
2020 IEEE Access  
Importantly, the parameters θ r is shared across domains, which can be trained by the transfer learning with excellent feature learning capability both for source and target domains [11] , [19] .  ...  graph convolutional network on the learned CNN features.  ... 
doi:10.1109/access.2020.3020149 fatcat:ppgxgdb7m5btlh4jz7y24ggrvi

Visual Transfer for Reinforcement Learning via Wasserstein Domain Confusion [article]

Josh Roy, George Konidaris
2020 arXiv   pre-print
WAPPO outperforms the prior state-of-the-art in visual transfer and successfully transfers policies across Visual Cartpole and two instantiations of 16 OpenAI Procgen environments.  ...  We introduce Wasserstein Adversarial Proximal Policy Optimization (WAPPO), a novel algorithm for visual transfer in Reinforcement Learning that explicitly learns to align the distributions of extracted  ...  Visual Cartpole and Procgen Easy are evaluated across 5 trials and Procgen Hard is evaluated across 3 trials. There is one source domain and one target domain per trial.  ... 
arXiv:2006.03465v1 fatcat:yfobpv4tdrhp5njilmgygf5qnu

Graph Adaptive Semantic Transfer for Cross-domain Sentiment Classification [article]

Kai Zhang, Qi Liu, Zhenya Huang, Mingyue Cheng, Kun Zhang, Mengdi Zhang, Wei Wu, Enhong Chen
2022 arXiv   pre-print
Cross-domain sentiment classification (CDSC) aims to use the transferable semantics learned from the source domain to predict the sentiment of reviews in the unlabeled target domain.  ...  Specifically, we present Graph Adaptive Semantic Transfer (GAST) model, an adaptive syntactic graph embedding method that is able to learn domain-invariant semantics from both word sequences and syntactic  ...  transferred across domains.  ... 
arXiv:2205.08772v1 fatcat:kgzhmx5y3bbxjnaragigbsd3tq

Transferability of Brain decoding using Graph Convolutional Networks [article]

Yu Zhang, Pierre Bellec
2020 bioRxiv   pre-print
The transferability of learned graph representations were evaluated under different circumstances, including knowledge transfer across cognitive domains, between different groups of subjects, and among  ...  Our results indicate that in contrast to natural images, the scanning condition, instead of task domain, has a larger impact on feature transfer for medical imaging.  ...  The transfer model achieved the highest decoding accuracy on the Motor task (92.5%) where the five types of body movements can be visually separated after the projection of learned graph representations  ... 
doi:10.1101/2020.06.21.163964 fatcat:334vgbbki5ccxajrrh7mdfa7a4

Graph Transfer Learning via Adversarial Domain Adaptation with Graph Convolution [article]

Quanyu Dai, Xiao-Ming Wu, Jiaren Xiao, Xiao Shen, Dan Wang
2022 arXiv   pre-print
To tackle this problem, we propose a novel graph transfer learning framework AdaGCN by leveraging the techniques of adversarial domain adaptation and graph convolution.  ...  Existing methods for single network learning cannot solve this problem due to the domain shift across networks.  ...  The Laplacian smoothing on node features with graph convolution in the representation learner enables an easy knowledge transfer across networks.  ... 
arXiv:1909.01541v3 fatcat:uspa64hdvnapbbxcntlorobdoi

Towards Cross-Domain Learning for Social Video Popularity Prediction

Suman Deb Roy, Tao Mei, Wenjun Zeng, Shipeng Li
2013 IEEE transactions on multimedia  
Index Terms-Cross-domain media retrieval, social media, transfer learning, Twitter, video popularity. 1520-9210 © 2013 IEEE  ...  We develop a transfer learning algorithm that can learn topics from social streams allowing us to model the social prominence of video content and improve popularity predictions in the video domain.  ...  The transfer graph presents a unified graph structure to represent the task of transfer learning from social domain to video domain.  ... 
doi:10.1109/tmm.2013.2265079 fatcat:jycldmgh7rcv7a52ii5ay7oz4i
« Previous Showing results 1 — 15 out of 49,624 results