Filters








4,398 Hits in 6.7 sec

Zhejiang University at TRECVID 2006

Yanan Liu, Fei Wu, Yueting Zhuang, Shengyi Zhou
2006 TREC Video Retrieval Evaluation  
SimFusion is an effective algorithm to reinforce or propagate the similarity relations between multi-modalities. LPP is an optimal combination of linear and nonlinear dimensionality reduction method.  ...  Although any uni-modality type partially expresses limited semantics less or more, video semantics are fully manifested only by interaction and integration of any unimodal.  ...  A multimodal analysis method for semantic understanding of video includes a fusion step to combine the results of several single media analysis.  ... 
dblp:conf/trecvid/Liu0ZZ06 fatcat:rigk4e4rzngbfjelztrttvm3a4

A New Approach to Cross-Modal Retrieval

Tingting Gou, Libo Liu, Qian Liu, Zhen Deng
2019 Journal of Physics, Conference Series  
The semantic correlation matching algorithm, which attempts to integrate semantic information based on the unified space obtained by CCA learning, achieves an improvement in the effect of cross-modal retrieval  ...  However, there is still a lot of room for the algorithm to make further progress in the generation of shared subspace.  ...  By employing two independent models of deep neural networks , the approach can map the low-level features of images and texts to their semantic subspace is proposed in [9] .  ... 
doi:10.1088/1742-6596/1288/1/012044 fatcat:ynevl6jjpfcere3nlwihls7cmq

Semi-Supervised Cross-Modal Retrieval Based on Discriminative Comapping

Li Liu, Xiao Dong, Tianshi Wang
2020 Complexity  
Most cross-modal retrieval methods based on subspace learning just focus on learning the projection matrices that map different modalities to a common subspace and pay less attention to the retrieval task  ...  In the process of projection matrix learning, a linear discriminant constraint is introduced to preserve the original class information in different modal spaces.  ...  : (1) e proposed joint formulation seamlessly combines semi-supervised learning, task-related learning, and linear discriminative analysis into a unified framework for cross-modal retrieval (2) e class  ... 
doi:10.1155/2020/1462429 fatcat:nvdtfct4crdozf5wa6fppdaaaq

Emotion recognition from scrambled facial images via many graph embedding

Richard Jiang, Anthony T.S. Ho, Ismahane Cheheb, Noor Al-Maadeed, Somaya Al-Maadeed, Ahmed Bouridane
2017 Pattern Recognition  
To handle with chaotic signals from face scrambling, in this paper, we propose an new approach -Many Graph Embedding (MGE) to discover discriminative patterns from the subspaces of chaotic patterns, where  ...  An immediate impact from face scrambling is that conventional semantic facial components become not identifiable, and 3D face models cannot be clearly fitted to a scrambled image.  ...  linear discriminant analysis (FLD) [30] .  ... 
doi:10.1016/j.patcog.2017.02.003 fatcat:bmwh4qo4zbd7hapgidktaa4xxu

Spectral regression

Deng Cai, Xiaofei He, Jiawei Han
2007 Proceedings of the 15th international conference on Multimedia - MULTIMEDIA '07  
Recently, several researchers have considered manifold ways to address this issue, such as Locality Preserving Projections, Augmented Relation Embedding, and Semantic Subspace Projection.  ...  In this paper, by using techniques from spectral graph embedding and regression, we propose a unified framework, called spectral regression, for learning an image subspace.  ...  Acknowledgments We would like to thank Xinjing Wang at MSRA for providing the data. The work was supported in part by the U.S.  ... 
doi:10.1145/1291233.1291329 dblp:conf/mm/CaiHH07 fatcat:yaoujh7kkfgy7jx5nu7bc3j2nm

Semantic Correlation based Deep Cross-Modal Hashing for Faster Retrieval

2019 VOLUME-8 ISSUE-10, AUGUST 2019, REGULAR ISSUE  
The multi-modal data has different statistical properties so there is a need to have method which finds semantic correlation between them.  ...  Due to growth of multi-modal data, large amount of data is being generated. Nearest Neighbor (NN) search is used to retrieve information but it suffers when there is high-dimensional data.  ...  To reduce really huge semantic gap between the representations of images and texts, [22] proposed a deep based visual semantic hashing (DVSH) approach where as a first step, they used deep learning to  ... 
doi:10.35940/ijitee.i8157.0881019 fatcat:jy2hfk7vx5dpzcr74lbvtvml7i

Face Recognition in the Scrambled Domain via Salience-Aware Ensembles of Many Kernels

Richard Jiang, Somaya Al-Maadeed, Ahmed Bouridane, Danny Crookes, M. Emre Celebi
2016 IEEE Transactions on Information Forensics and Security  
While chaotic pattern recognition is still a challenging task, in this paper we propose a new ensemble approach -Many-Kernel Random Discriminant Analysis (MK-RDA) to discover discriminative patterns from  ...  We also incorporate a salience-aware strategy into the proposed ensemble method to handle chaotic facial patterns in the scrambled domain, where random selections of features are made on semantic components  ...  Though semantic approaches have attained great success in facial analysis, they need a robust scheme to map a 2D image into its semantic feature space or 3D deformable model.  ... 
doi:10.1109/tifs.2016.2555792 fatcat:d2gkr67vszebjb3xntnqr2dtge

Modality Mixture Projections for Semantic Video Event Detection

Jialie Shen, Dacheng Tao, Xuelong Li
2008 IEEE transactions on circuits and systems for video technology (Print)  
In this paper, we use a subspace selection technique to achieve fast and accurate video event detection using a subspace selection technique.  ...  Furthermore, the training stage is carried out once and we have a unified transformation matrix to project different modalities.  ...  With MMP, all the raw video feature data from various heterogenous information source can be effectively fused to obtain single unified subspace via learning process.  ... 
doi:10.1109/tcsvt.2008.2005607 fatcat:zjjj3pmtvnagldgcz3aqthnfcm

Modality-dependent Cross-media Retrieval [article]

Yunchao Wei, Yao Zhao, Zhenfeng Zhu, Shikui Wei, Yanhui Xiao, Jiashi Feng, Shuicheng Yan
2015 arXiv   pre-print
Specifically, by jointly optimizing the correlation between images and text and the linear regression from one modal space (image or text) to the semantic space, two couples of mappings are learned to  ...  project images and text from their original feature spaces into two common latent subspaces (one for I2T and the other for T2I).  ...  L(V, W ) is a linear regression term from one modal feature space (image or text) to the semantic space, used to centralize the multi-modal data with the same semantics in the common latent subspace.  ... 
arXiv:1506.06628v2 fatcat:vbnedfrefzdldk67prpojzl2v4

Latent Fisher Discriminant Analysis [article]

Gang Chen
2013 arXiv   pre-print
We test our method on MUSK and Corel data sets and yield competitive results compared to the baseline approach.  ...  Linear Discriminant Analysis (LDA) is a well-known method for dimensionality reduction and classification. Previous studies have also extended the binary-class case into multi-classes.  ...  To deal with non-linear scenarios, the kernel approach [21] can be applied easily via the so-called kernel trick to extend LDA to its kernel version, called kernel discriminant analysis [2] , that can  ... 
arXiv:1309.5427v1 fatcat:ovxfl3yizjavzdsuxgjwmpmkj4

Cross-Modal Subspace Learning with Scheduled Adaptive Margin Constraints

David Semedo, Joao Magalhaes
2019 Proceedings of the 27th ACM International Conference on Multimedia - MM '19  
Cross-modal embeddings, between textual and visual modalities, aim to organise multimodal instances by their semantic correlations.  ...  State-of-the-art approaches use maximum-margin methods, based on the hinge-loss, to enforce a constant margin m, to separate projections of multimodal instances from different categories.  ...  Deep Canonical Correlation Analysis (DCCA) was adopted in [33] to match images and text, using non-linear projections.  ... 
doi:10.1145/3343031.3351030 dblp:conf/mm/SemedoM19 fatcat:mbrtmkfvxrcc5nplijthor2n2e

A Cross-Media Retrieval Method Based on Semisupervised Learning and Alternate Optimization

Junzheng Li, Wei Zhu, Yanchun Yang, Xiyuan Zheng
2021 Mobile Information Systems  
Simultaneously, we combine the linear regression term, correlation analysis term, and feature selection term into a joint cross-media learning framework.  ...  The most difficult task for cross-media retrieval lies in the potential correlation between different modalities data and how to overcome the semantic gap.  ...  Typical methods are Canonical Correlation Analysis (CCA) [15] , Semantic Matching (SM) [16] , Semantic Relevance Matching (SCM) [17] , T-V CCA [18] , Generalized Multiview Analysis Linear Discriminant  ... 
doi:10.1155/2021/9947644 doaj:90a40b8f33a34c6fa8e02d33c8713613 fatcat:3hfjrkeerfa3tgbtlmwd3o6qd4

Fast Updating Algorithms for Latent Semantic Indexing

Eugene Vecharynski, Yousef Saad
2014 SIAM Journal on Matrix Analysis and Applications  
Retrieve information (text documents) relevant to a given query. Text data are non-static, i.e., constantly updated. Consider the Latent Semantic Indexing (LSI) approach for information retrieval.  ...  Develop a unified framework for the SVD updating in LSI.  ...  The left search subspace is compressed using the SVD of (I − U k U T k )D. The SVD of H D now costs O((k + l) 2 (k + p)), which is linear in p.  ... 
doi:10.1137/130940414 fatcat:tcf52gmnwrejnjaujwhuymfv64

Cross-Media Retrieval via Semantic Entity Projection [chapter]

Lei Huang, Yuxin Peng
2016 Lecture Notes in Computer Science  
To address this challenging problem, most existing approaches project heterogeneous features into a unified feature space to facilitate their similarity computation.  ...  By considering the above issues, we propose a new approach to cross-media retrieval via semantic entity projection (SEP) in this paper. Our approach consists of three main steps.  ...  They utilize unsupervised methods like CCA to obtain a unified subspace, but there exists no explicit semantic meanings in this unified subspace.  ... 
doi:10.1007/978-3-319-27671-7_23 fatcat:zhy4dmfgarho5oe2pwfahmvudu

Semantic convex matrix factorisation for cross-media retrieval

Yixian Fang, Yuwei Ren, Huaxiang Zhang
2019 IET Image Processing  
To address these problems, a semantic convex matrix factorisation subspace learning approach is proposed for cross-media retrieval between image and text.  ...  In addition, many presented approaches directly mapped different modalities into an isomorphic semantic space to conduct the similarity measurement of different modalities, which also resulted in the loss  ...  For above reasons, a semantic convex matrix factorisation (SCMF) subspace learning approach is proposed to enhance the discriminability of matrix factorisation by semantic information.  ... 
doi:10.1049/iet-ipr.2018.5853 fatcat:sdtyhoiowbapvhwdwya6qvipt4
« Previous Showing results 1 — 15 out of 4,398 results