Filters








2,046 Hits in 5.5 sec

Image Retrieval with Structured Object Queries Using Latent Ranking SVM [chapter]

Tian Lan, Weilong Yang, Yang Wang, Greg Mori
2012 Lecture Notes in Computer Science  
Our learning method is an extension of the ranking SVM with latent variables, which we call latent ranking SVM.  ...  We consider image retrieval with structured object queriesqueries that specify the objects that should be present in the scene, and their spatial relations.  ...  The inference takes around 0.05 sec per image in MATLAB on a 2.8GHZ CPU 8GB RAM PC. Learning with Latent Ranking SVM We learn our model in a latent ranking SVM framework.  ... 
doi:10.1007/978-3-642-33783-3_10 fatcat:f4u2aee765b6jma5ukrnqz2ed4

Ranking and retrieval of image sequences from multiple paragraph queries

Gunhee Kim, Seungwhan Moon, Leonid Sigal
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
, based upon the structural SVM with latent variables Retrieval: Find optimal image sequence for a given wrt Weights to learn Feature vectors Feature vectors One-to-one compatibility of  ...  sequence.• 80% of blog posts used as the training set, and the others as test set • For a test blog, Quantitative comparison btw fixed models vs latent structural SVM • Five different text segmentation  ... 
doi:10.1109/cvpr.2015.7298810 dblp:conf/cvpr/KimMS15 fatcat:yktzyav6uvawbkyg52kwb6uxvy

A low rank structural large margin method for cross-modal ranking

Xinyan Lu, Fei Wu, Siliang Tang, Zhongfei Zhang, Xiaofei He, Yueting Zhuang
2013 Proceedings of the 36th international ACM SIGIR conference on Research and development in information retrieval - SIGIR '13  
The latent lowrank embedding space is discriminatively learned by structural large margin learning to optimize for certain ranking criteria directly.  ...  , which we call Latent Semantic Cross-Modal Ranking (LSCMR).  ...  Since we assume that the query texts and the target images are embedded into a common latent space, respectively, LSCMR adapts the original structural SVM to learn the optimal U * and V * which maximize  ... 
doi:10.1145/2484028.2484039 dblp:conf/sigir/LuWTZHZ13 fatcat:2cn4fjzz4vhizck2tok5u6tj3u

A Latent Model for Visual Disambiguation of Keyword-based Image Search

Kong-Wah Wan, Ah-Hwee Tan, Joo-Hwee Lim, Liang-Tien Chia, Sujoy Roy
2009 Procedings of the British Machine Vision Conference 2009  
A sense-specific image classifier is then learnt by combining information from the latent visual sense model, and used to cluster and re-rank the polysemous images from the original query keyword into  ...  Given a query keyword and the images retrieved by issuing the query to an image search engine, we first learn a latent visual sense model of these polysemous images.  ...  Given that we are also retrieving images using sense-specific queries, an obvious approach is to bootstrap sense-specific classifiers from these images. We shall call this method Sense-Specific SVM.  ... 
doi:10.5244/c.23.67 dblp:conf/bmvc/WanTLCR09 fatcat:2pgi4flyvjfztgwirppx2zqclm

Cross-media semantic representation via bi-directional learning to rank

Fei Wu, Xinyan Lu, Zhongfei Zhang, Shuicheng Yan, Yong Rui, Yueting Zhuang
2013 Proceedings of the 21st ACM international conference on Multimedia - MM '13  
, which do not make full use of bi-directional ranking examples (bi-directional ranking means that both text-query-image and image-querytext ranking examples are utilized in the training period) to achieve  ...  The latent space embedding is discriminatively learned by the structural large margin learning for optimization with certain ranking criteria (mean average precision in this paper) directly.  ...  Similarly, consider the other direction of the retrieval, i.e., ranking text documents from image queries. To apply structural SVM, the process is analogous.  ... 
doi:10.1145/2502081.2502097 dblp:conf/mm/WuLZYRZ13 fatcat:ni7x2naeavcgdalyqnrg4ug464

Learning a Multi-Concept Video Retrieval Model with Multiple Latent Variables

Amir Mazaheri, Boqing Gong, Mubarak Shah
2018 ACM Transactions on Multimedia Computing, Communications, and Applications (TOMCCAP)  
Our approach, which can be considered as a latent ranking SVM, integrates the advantages of various recent works on text and image retrieval, such as choosing ranking over structured prediction and modeling  ...  Effective and efficient video retrieval has become a pressing need in the "big video" era and how to deal with multi-concept queries is a central component.  ...  Our approach, which can be considered as a latent ranking SVM [10] , integrates different advantages of the recent works on text and multi-attribute based image retrieval.  ... 
doi:10.1145/3176647 fatcat:hbj5ssfgfjffjnjiqvwggowotm

Learning a Multi-concept Video Retrieval Model with Multiple Latent Variables

Amir Mazaheri, Boqing Gong, Mubarak Shah
2016 2016 IEEE International Symposium on Multimedia (ISM)  
Our approach, which can be considered as a latent ranking SVM, integrates the advantages of various recent works on text and image retrieval, such as choosing ranking over structured prediction and modeling  ...  Effective and efficient video retrieval has become a pressing need in the "big video" era and how to deal with multi-concept queries is a central component.  ...  Our approach, which can be considered as a latent ranking SVM [10] , integrates different advantages of the recent works on text and multi-attribute based image retrieval.  ... 
doi:10.1109/ism.2016.0132 dblp:conf/ism/MazaheriGS16 fatcat:mjuacchiv5h3lauentz6mpbfuy

Content-based Video Indexing and Retrieval Using Corr-LDA [article]

Rahul Radhakrishnan Iyer, Sanjeel Parekh, Vikas Mohandoss, Anush Ramsurat, Bhiksha Raj, Rita Singh
2019 arXiv   pre-print
We use the concept-level matching provided by corr-LDA to build correspondences between text and multimedia, with the objective of retrieving content with increased accuracy.  ...  We present our work on Content-based Video Indexing and Retrieval using the Correspondence-Latent Dirichlet Allocation (corr-LDA) probabilistic framework.  ...  Rank SVM based approach is proposed in where linear classifiers are trained for an image with relevant and irrelavant captions.  ... 
arXiv:1602.08581v2 fatcat:3me6ugirsjaadpx2cubwcsxe3m

Topic Correlations for Cross-Modal Multimedia Information Retrieval [chapter]
English

Jing Yu, Zengchang Qin
2015 Gate to Computer Science and Research  
The latent semantic relations between texts and images can be reflected by correlations between the word topics and topics of image features.  ...  In this paper, we propose a novel approach for crossmodal multimedia retrieval by jointly modeling the text and image components of multimedia documents.  ...  The observable images and texts can be used to estimate hyper-parameters or the latent structure of the model.  ... 
doi:10.15579/gcsr.vol4.ch3 fatcat:flmmlt46rferrexiodywjb7sx4

A New Algorithm for Sketch-Based Fashion Image Retrieval Based on Cross-Domain Transformation

Haopeng Lei, Simin Chen, Mingwen Wang, Xiangjian He, Wenjing Jia, Sibo Li, Amr Tolba
2021 Wireless Communications and Mobile Computing  
However, the current mainstream retrieval methods are still limited to using text or exemplar images as input.  ...  Specifically, when retrieving on our Fashion Image dataset, the accuracy of our model ranks the correct match at the top-1 which is 96.6%, 92.1%, 91.0%, and 90.5% for clothes, pants, skirts, and shoes,  ...  On this dataset, we compare our model with BoW-HOG + rank-SVM, Improved Sketch-a-Net (ISN), Dense-HOG + Rank-SVM, and 3D shape (3DS).  ... 
doi:10.1155/2021/5577735 fatcat:izmg3ezh7bf7dlxvcxmgpkvb3m

Click-through-based cross-view learning for image search

Yingwei Pan, Ting Yao, Tao Mei, Houqiang Li, Chong-Wah Ngo, Yong Rui
2014 Proceedings of the 37th international ACM SIGIR conference on Research & development in information retrieval - SIGIR '14  
One of the fundamental problems in image search is to rank image documents according to a given textual query.  ...  in the latent subspace and preserving the inherent structure in each original space.  ...  For any query, sorting by its corresponding values for all its associated images gives the retrieval ranking for these images. The algorithm is given in Algorithm 1.  ... 
doi:10.1145/2600428.2609568 dblp:conf/sigir/PanYMLNR14 fatcat:677byckxt5gihd6z4s3epm3bva

Discriminative multi-view Privileged Information learning for image re-ranking [article]

Jun Li, Chang Xu, Wankou Yang, Changyin Sun, Dacheng Tao, Hong Zhang
2018 arXiv   pre-print
a unified training framework for generating the latent subspaces with sufficient discriminating power.  ...  Due to the inconsistency in the visual appearance, this practice tends to degrade the retrieval accuracy particularly when the image ROI, which is usually interpreted as the image objectness, accounts  ...  The resulting PI-aware subspace preserves sufficient discrimination in the image, and thus can be used for generating discriminative objectness-aware latent representation for accurate re-ranking.  ... 
arXiv:1808.04437v1 fatcat:exveffacpvhgfbmcmts7jzwsv4

Comparative Study of Some Supervised Machine Learning Algorithms for Information Retrieval

Kissinger Sunday, Muhammad Bello Aliyu
2020 Saudi Journal of Engineering and Technology  
The selected algorithms were critically studied in line with the available matching models for information retrieval.  ...  Models like the Vector space model, Binary model, probabilistic models, Inverted Index, Latent semantic Analysis and the Latent Semantic Index models were respectively examined.  ...  In order to rank SVM for IR, we generate each instance x from a query-document pair and followed by a label with one rank (from the 2 possible ranks defined above).  ... 
doi:10.36348/sjet.2020.v05i03.003 fatcat:golmm7anbrcxbpuhslf6bfhel4

A Max-Margin Riffled Independence Model for Image Tag Ranking

Tian Lan, Greg Mori
2013 2013 IEEE Conference on Computer Vision and Pattern Recognition  
We propose Max-Margin Riffled Independence Model (MMRIM), a new method for image tag ranking modeling the structured preferences among tags.  ...  Our model integrates the max-margin formalism with riffled independence factorizations proposed in [10], which naturally allows for structured learning and efficient ranking.  ...  Image Retrieval We use all test images as queries and all training images as the database.  ... 
doi:10.1109/cvpr.2013.399 dblp:conf/cvpr/LanM13 fatcat:h4t4j7nq3zbe5fkjrvx33ukxze

Object image retrieval by exploiting online knowledge resources

Gang Wang, David Forsyth
2008 2008 IEEE Conference on Computer Vision and Pattern Recognition  
We describe a method to retrieve images found on web pages with specified object class labels, using an analysis of text around the image and of image appearance.  ...  Our method determines whether an object is both described in text and appears in a image using a discriminative image model and a generative text model.  ...  For example, we use "frog amphibian" to extract frog images. Approach Our goal is to retrieve object images from noisy web page with image and text cues.  ... 
doi:10.1109/cvpr.2008.4587818 dblp:conf/cvpr/WangF08 fatcat:42syv26jebbwflm47nd57p3fiu
« Previous Showing results 1 — 15 out of 2,046 results