Filters








6,174 Hits in 5.6 sec

Intrinsic Geometric Information Transfer Learning on Multiple Graph-Structured Datasets [article]

Jaekoo Lee, Hyunjae Kim, Jongsun Lee, Sungroh Yoon
2016 arXiv   pre-print
By transferring the intrinsic geometric information learned in the source domain, our approach can help us to construct a model for a new but related task in the target domain without collecting new data  ...  In this paper, we attempt to advance deep learning for graph-structured data by incorporating another component, transfer learning.  ...  In the context of graphs, we call the transferred information the intrinsic geometric information.  ... 
arXiv:1611.04687v2 fatcat:62c4y4gysncdhnknka7jtl4fdm

Learning Geometrically Disentangled Representations of Protein Folding Simulations [article]

N. Joseph Tatro, Payel Das, Pin-Yu Chen, Vijil Chenthamarakshan, Rongjie Lai
2022 arXiv   pre-print
Specifically, we present a geometric autoencoder framework to learn separate latent space encodings of the intrinsic and extrinsic geometries of the protein structure.  ...  This work focuses on learning a generative neural network on a structural ensemble of a drug-target protein, e.g.  ...  the test dataset after transferring a trained model to other protein structures.  ... 
arXiv:2205.10423v1 fatcat:2riqmesmefc2pj646ksyxdeuzi

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)  
multiple datasets via Inter-Graph Transfer.  ...  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  ...  We first learn and propagate compact high-level semantic graph representation within one dataset via Intra-Graph Reasoning, and then transfer and fuse the semantic information across multiple datasets  ... 
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
multiple datasets via Inter-Graph Transfer.  ...  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  ...  We first learn and propagate compact high-level semantic graph representation within one dataset via Intra-Graph Reasoning, and then transfer and fuse the semantic information across multiple datasets  ... 
arXiv:1904.04536v1 fatcat:di2yce3ytbhadml5lljt7yn66m

Artificial Bee Colony Based Multiview Clustering ABC MVC for Graph Structure Fusion in Benchmark Datasets

N. Kamalraj
2020 Zenodo  
These algorithms don't consider the correlation of graph structure between multiple views, and the clustering results highly based on the value of predefined affinity graphs.  ...  ABC MVC model is based on the presumption with the purpose of intrinsic underlying graph structure would assign related connected part in each graph toward the similar group.  ...  To detect the intrinsic structure shared via varied views, the Hadamard product is utilized in order to conserve the general edges in multiple graphs.  ... 
doi:10.5281/zenodo.3843242 fatcat:ggrwzutvozbrdpzig3qfeaxu3q

Semi-Supervised Tensor-Based Graph Embedding Learning and Its Application to Visual Discriminant Tracking

Weiming Hu, Jin Gao, Junliang Xing, Chao Zhang, Stephen Maybank
2017 IEEE Transactions on Pattern Analysis and Machine Intelligence  
We design two graphs for characterizing the intrinsic local geometrical structure of the tensor samples of the object and the background.  ...  information obtained from earlier times is transferred.  ...  The intrinsic local geometrical and discriminative structures of the tensor samples are effectively represented.  We propose a transfer-learning-based semi-supervised learning method to adjust the 2-order  ... 
doi:10.1109/tpami.2016.2539944 pmid:26978551 fatcat:g6qus5mnrbfzphw4ubgwnuhayq

AN EFFICIENT REPRESENTATION OF 3D BUILDINGS: APPLICATION TO THE EVALUATION OF CITY MODELS

O. Ennafii, A. Le Bris, F. Lafarge, C. Mallet
2021 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
They are based on graph kernels and Scattering Network. They are here evaluated in the challenging framework of quality evaluation of building models.  ...  In particular, for them to be used in supervised learning tasks, such a representation should be scalable and transferable to various environments as only a few reference training instances would be available  ...  Geometric features make use of graph kernels (cf. Section 2.2) for geometric features and ScatNets (cf. Section 2.3) for grid structures based ones.  ... 
doi:10.5194/isprs-archives-xliii-b2-2021-329-2021 fatcat:cuqaeqzln5hvpmi66xd6kjr22e

Intrinsic-Extrinsic Preserved GANs for Unsupervised 3D Pose Transfer [article]

Haoyu Chen, Hao Tang, Henglin Shi, Wei Peng, Nicu Sebe, Guoying Zhao
2021 arXiv   pre-print
The extensive experiments on various 3D datasets of humans, animals and hands qualitatively and quantitatively demonstrate the generality of our approach.  ...  Specifically, a novel Intrinsic-Extrinsic Preserved Generative Adversarial Network (IEP-GAN) is presented for both intrinsic (i.e., shape) and extrinsic (i.e., pose) information preservation.  ...  computations, which makes them unsuitable for learning on largescale datasets.  ... 
arXiv:2108.07520v2 fatcat:ephakoeqd5aprpwdrhb6mbcyia

A Discriminative-Based Geometric Deep Learning Model for Cross Domain Recommender Systems

John Kingsley Arthur, Conghua Zhou, Eric Appiah Mantey, Jeremiah Osei-Kwakye, Yaru Chen
2022 Applied Sciences  
Traditional machine learning and deep learning methods are not designed to learn from complex data representations such as graphs, manifolds and 3D objects.  ...  and visual information from the structure of the recommended 3D objects.  ...  In this approach, a local representation learning is introduced into the structure of deep learning techniques to effectively capture the local geometric and visual information from the structure of the  ... 
doi:10.3390/app12105202 fatcat:u4mc7e5xdrfjxdygd5b55vr3ue

Discriminative Transfer Learning on Manifold [chapter]

Zheng Fang, Zhongfei (Mark) Zhang
2013 Proceedings of the 2013 SIAM International Conference on Data Mining  
Empirical study on benchmark datasets validates that DTLM improves the classification accuracy consistently compared with the state-of-theart transfer learning methods.  ...  Collective matrix factorization has achieved a remarkable success in document classification in the literature of transfer learning.  ...  Acknowledgment This work is supported in part by National Basic Research Program of China (2012CB316400), ZJU-Alibaba Financial Joint Lab, and Zhejiang Provincial Engineering Center on Media Data Cloud  ... 
doi:10.1137/1.9781611972832.60 dblp:conf/sdm/FangZ13 fatcat:bp3ubjy2brhjfnokbpfug2yvxq

Recapture as You Want [article]

Chen Gao, Si Liu, Ran He, Shuicheng Yan, Bo Li
2020 arXiv   pre-print
It effectively addresses the non-rigid deformation issue and well preserves the intrinsic structure/appearance with rich texture details.  ...  LGR module utilizes body skeleton knowledge to construct a layout graph that connects all relevant part features, where graph reasoning mechanism is used to propagate information among part nodes to mine  ...  Semantic-aware Geometric Transformation The semantic-aware geometric transformation network is shown in Fig. 3 , which aims to learn the mapping S A → S B for producingŜ B conditioned on K B .  ... 
arXiv:2006.01435v1 fatcat:3wvfhxm5r5cgnezxigfr7qzkru

Exploiting Relational Information in Social Networks using Geometric Deep Learning on Hypergraphs

Devanshu Arya, Marcel Worring
2018 Proceedings of the 2018 ACM on International Conference on Multimedia Retrieval - ICMR '18  
However, there are two challenges: (a) a social network has an intrinsic community structure.  ...  We study the behavior of our method by performing experiments on CLEF dataset consisting of images from Flickr, an online photo sharing social network.  ...  In this work, we focus on applying convolution network on graphs in order to learn the intrinsic relations in social networks.  ... 
doi:10.1145/3206025.3206062 dblp:conf/mir/AryaW18 fatcat:efr5s4otpvc3hdcdgmxv6gu7my

Self-supervised Learning with Geometric Constraints in Monocular Video: Connecting Flow, Depth, and Camera [article]

Yuhua Chen, Cordelia Schmid, Cristian Sminchisescu
2019 arXiv   pre-print
We also show good generalization for transfer learning in YouTube videos.  ...  We demonstrate the effectiveness of our method on KITTI and Cityscapes, where we outperform previous self-supervised approaches on multiple tasks.  ...  We additionally train the framework on Cityscapes [7] , and study how well the proposed models transfer across datasets.  ... 
arXiv:1907.05820v2 fatcat:6ithess5zbdwlcqjga3ko25eru

Robust Structure Preserving Nonnegative Matrix Factorization for Dimensionality Reduction

Bingfeng Li, Yandong Tang, Zhi Han
2016 Mathematical Problems in Engineering  
In RSPNMF, a local affinity graph and a distant repulsion graph are constructed to encode the geometrical information, and noisy data influence is alleviated by characterizing the data reconstruction term  ...  However, there are still two major drawbacks for NMF: (a) NMF can only perform semantic factorization in Euclidean space, and it fails to discover the intrinsic geometrical structure of high-dimensional  ...  In graph embedding algorithm, graph characterizes the prior knowledge of geometric structure from data distribution.  ... 
doi:10.1155/2016/7474839 fatcat:qh7vpl6jmvb4bjrdpubqreukpa

SpiralNet++: A Fast and Highly Efficient Mesh Convolution Operator [article]

Shunwang Gong, Lei Chen, Michael Bronstein, Stefanos Zafeiriou
2019 arXiv   pre-print
We explicitly formulate the order of aggregating neighboring vertices, instead of learning weights between nodes, and then a fully connected layer follows to fuse local geometric structure information  ...  Intrinsic graph convolution operators with differentiable kernel functions play a crucial role in analyzing 3D shape meshes.  ...  This method allows computational models that are composed of multiple layers to learn representations of irregular data structures, such as graphs and meshes.  ... 
arXiv:1911.05856v1 fatcat:4hjk7nehqjdgffykdpxc6ek7xi
« Previous Showing results 1 — 15 out of 6,174 results