562 Hits in 5.0 sec

Cross-domain correspondence for sketch-based 3D model retrieval using convolutional neural network and manifold ranking

Shichao Jiao, Xie Han, Fengguang Xiong, Fusheng Sun, Rong Zhao, Liqun Kuang
2020 IEEE Access  
To address these problems, we propose cross-domain correspondence method for sketch-based 3D model retrieval based on manifold ranking.  ...  INDEX TERMS Sketch, 3D model retrieval, deep learning, semantic labels, manifold ranking, convolutional neural network.  ...  CONCLUSION In this paper, we proposed a query-by-sketch method for 3D model retrieval based on deep learning and manifold ranking.  ... 
doi:10.1109/access.2020.3006585 fatcat:mr5hempqoff4zegqasz73qvor4

Siamese CNN-BiLSTM Architecture for 3D Shape Representation Learning

Guoxian Dai, Jin Xie, Yi Fang
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
Learning a 3D shape representation from a collection of its rendered 2D images has been extensively studied.  ...  In this paper, by employing recurrent neural network to efficiently capture features across different views, we propose a siamese CNN-BiLSTM network for 3D shape representation learning.  ...  for 3D shape representation learning.  ... 
doi:10.24963/ijcai.2018/93 dblp:conf/ijcai/DaiXF18 fatcat:es6sey4wk5a7hlxdiovakgt25q

Thin-Slicing for Pose: Learning to Understand Pose without Explicit Pose Estimation

Suha Kwak, Minsu Cho, Ivan Laptev
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
We demonstrate the efficacy of the embedding on pose-based image retrieval and action recognition problems.  ...  This architecture allows us to learn a robust representation that captures differences in human poses by effectively factoring out variations in clothing, background, and imaging conditions in the wild  ...  We thank Greg Mori and Vadim Kantorov for fruitful discussions. This work was supported in part by Google Research Award and the ERC grants Activia and VideoWorld.  ... 
doi:10.1109/cvpr.2016.534 dblp:conf/cvpr/KwakCL16 fatcat:mve2ynmwg5g2hhqtlcdmyhqwp4

Cross‐modal semantic correlation learning by Bi‐CNN network

Chaoyi Wang, Liang Li, Chenggang Yan, Zhan Wang, Yaoqi Sun, Jiyong Zhang
2021 IET Image Processing  
Cross modal retrieval can retrieve images through a text query and vice versa. In recent years, cross modal retrieval has attracted extensive attention.  ...  Meanwhile, we enhance the semantic manifold by constructing cross modal ranking and within-modal discriminant loss to improve the division of semantic representation.  ...  Specifically, the deep CNN (i.e., ResNet) is applied for images representation and the shallow multisize kernels CNN generates muti-level texts feature.  ... 
doi:10.1049/ipr2.12176 fatcat:xbw2qc5tffdu7g3it7g6n4ljje

Camera Relocalization by Computing Pairwise Relative Poses Using Convolutional Neural Network

Zakaria Laskar, Iaroslav Melekhov, Surya Kalia, Juho Kannala
2017 2017 IEEE International Conference on Computer Vision Workshops (ICCVW)  
Chapter 3 concentrates on the problems of image retrieval and presents approaches for improving retrieval performance based on global representations and local spatial re-ranking strategies.  ...  By making this next stage in the pipeline differentiable the structure in data manifold can be exploited to learn better representations and pose estimates.  ... 
doi:10.1109/iccvw.2017.113 dblp:conf/iccvw/LaskarMKK17 fatcat:qhms7distnf2zprw3r6oqeouqa

Efficient Diffusion on Region Manifolds: Recovering Small Objects with Compact CNN Representations [article]

Ahmet Iscen, Giorgos Tolias, Yannis Avrithis, Teddy Furon, Ondrej Chum
2019 arXiv   pre-print
Query expansion is a popular method to improve the quality of image retrieval with both conventional and CNN representations. It has been so far limited to global image similarity.  ...  Small objects have been a common failure case of CNN-based retrieval.  ...  The Tesla K40 used for this research was donated by the NVIDIA Corporation.  ... 
arXiv:1611.05113v3 fatcat:cxyn3wszn5gltmqlmvrzfnfjhq

DeepDiffusion: Unsupervised Learning of Retrieval-adapted Representations via Diffusion-based Ranking on Latent Feature Manifold [article]

Takahiko Furuya, Ryutarou Ohbuchi
2021 arXiv   pre-print
Experimental evaluation using 3D shapes and 2D images demonstrates versatility as well as high accuracy of the DD algorithm. Code is available at  ...  Unsupervised learning of feature representations is a challenging yet important problem for analyzing a large collection of multimedia data that do not have semantic labels.  ...  by using scenarios of 3D shape retrieval and 2D image retrieval.  ... 
arXiv:2112.07082v1 fatcat:mc75afshjzdwpmwb5lzguwracy

Iterative Manifold Embedding Layer Learned by Incomplete Data for Large-scale Image Retrieval [article]

Jian Xu, Chunheng Wang, Chengzuo Qi, Cunzhao Shi, Baihua Xiao
2018 arXiv   pre-print
Existing manifold learning methods are not appropriate for image retrieval task, because most of them are unable to process query image and they have much additional computational cost especially for large  ...  We experiment on five public standard datasets for image retrieval. The proposed IME layer significantly outperforms related dimension reduction methods and manifold learning methods.  ...  The authors would like to thank the Associate Editor and the anonymous reviewers for their contributions to improve the quality of this paper.  ... 
arXiv:1707.09862v2 fatcat:7crk6zmtpzautb5vbaeikspvme

A Survey on Deep Visual Place Recognition

Carlo Masone, Barbara Caputo
2021 IEEE Access  
As mentioned earlier, representation learning via ranking losses requires to select positive and negative examples for each training image.  ...  LEARNING TO RANK Image retrieval is akin to a "learning-to-rank" problem, therefore it naturally lends itself to metric learning, i.e., learning image descriptors that represent well the similarity under  ... 
doi:10.1109/access.2021.3054937 fatcat:hc5fp2z4g5fldl7imkt7ixq4z4

UniformFace: Learning Deep Equidistributed Representation for Face Recognition

Yueqi Duan, Jiwen Lu, Jie Zhou
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
In this paper, we propose a new supervision objective named uniform loss to learn deep equidistributed representations for face recognition.  ...  Most existing methods aim to learn discriminative face features, encouraging large interclass distances and small intra-class variations.  ...  Implementation Details Detailed Setup of CNN: We utilized MXNet package [6] through the experiments and employed ResNet [13] as the CNN architecture for all the datasets.  ... 
doi:10.1109/cvpr.2019.00353 dblp:conf/cvpr/DuanL019 fatcat:s5uaxikpfnh3lkmvoilfzlcck4

SPNet: Deep 3D Object Classification and Retrieval using Stereographic Projection [article]

Mohsen Yavartanoo, Eu Young Kim, Kyoung Mu Lee
2019 arXiv   pre-print
We propose an efficient Stereographic Projection Neural Network (SPNet) for learning representations of 3D objects.  ...  Despite its lightness, the experiments on 3D object classification and shape retrievals demonstrate the high performance of the proposed method.  ...  Shi et al. in DeepPano [28] , projected each 3D shape into a panoramic view around its principal axis and used a CNN for learning the representations from these views.  ... 
arXiv:1811.01571v2 fatcat:mxcbfvozgbdj3mxqnws5sq7oyq

Efficient Diffusion on Region Manifolds: Recovering Small Objects with Compact CNN Representations

Ahmet Iscen, Giorgos Tolias, Yannis Avrithis, Teddy Furon, Ondrej Chum
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
Query expansion is a popular method to improve the quality of image retrieval with both conventional and CNN representations. It has been so far limited to global image similarity.  ...  Small objects have been a common failure case of CNN-based retrieval.  ...  The Tesla K40 used for this research was donated by the NVIDIA Corporation.  ... 
doi:10.1109/cvpr.2017.105 dblp:conf/cvpr/IscenTAFC17 fatcat:3n246p5dnzeujf44cknb5xegki

Deep Cross-modality Adaptation via Semantics Preserving Adversarial Learning for Sketch-based 3D Shape Retrieval [article]

Jiaxin Chen, Yi Fang
2018 arXiv   pre-print
Due to the large cross-modality discrepancy between 2D sketches and 3D shapes, retrieving 3D shapes by sketches is a significantly challenging task.  ...  3D shapes.  ...  The CNN and metric networks for each single modality (i.e., 2D sketches or 3D shapes) is trained by importance-aware metric learning through mining the hardest training samples.  ... 
arXiv:1807.01806v1 fatcat:xuww7nfkkncqdbszqcgf7sls3e

Learning Discriminative 3D Shape Representations by View Discerning Networks [article]

Biao Leng, Cheng Zhang, Xiaocheng Zhou, Cheng Xu, Kai Xu
2018 arXiv   pre-print
In view-based 3D shape recognition, extracting discriminative visual representation of 3D shapes from projected images is considered the core problem.  ...  To resolve this problem, we propose a novel deep neural network, View Discerning Network, which learns to judge the quality of views and adjust their contributions to the representation of shapes.  ...  Thus, we conclude that 10 images for one shape are enough for 3D shape retrieval task. And in our further experiment, we will exert 10 images from different angles to represent the 3D shapes.  ... 
arXiv:1808.03823v2 fatcat:qfx5a67ikrgwdoli2tgazs2iy4

A Review on Deep Learning Approaches for 3D Data Representations in Retrieval and Classifications

Abubakar Sulaiman Gezawa, Yan Zhang, Qicong Wang, Lei Yunqi
2020 IEEE Access  
Finally, some possible directions for future researches were suggested. INDEX TERMS 3D data representation, 3D deep learning, 3D models dataset, computer vision, classification, retrieval.  ...  Deep learning approach has been used extensively in image analysis tasks.  ...  In [112] , Sinha et al. convert a 3D object into a geometry image and use CNNs to learn 3D shapes.  ... 
doi:10.1109/access.2020.2982196 fatcat:jnya5rscynf3zm7efuucqxafri
« Previous Showing results 1 — 15 out of 562 results