623,469 Hits in 4.9 sec

Image Re-ranking using Information Gain and Relative Consistency through Multi-graph Learning

Poonam N., Supriya A.
2016 International Journal of Computer Applications  
Our submodular re-ranking framework can be easily used in re-ranking problems for realtime search engines.  ...  We deal with mainly two terms: information gain and relative ranking consistency among multiple modalities.  ...  To improve the re-ranking performance we use the initial ranks of retrieved images from different feature channels as an extra information.  ... 
doi:10.5120/ijca2016911131 fatcat:cenvennv2repxgtpsikmwsgjse

Imcube @ MediaEval 2015 Retrieving Diverse Social Images Task: Multimodal Filtering and Re-ranking

Sebastian Schmiedeke, Pascal Kelm, Lutz Goldmann
2015 MediaEval Benchmarking Initiative for Multimedia Evaluation  
Therefore, we propose a multi-modal approach for filtering and re-ranking in order to improve the relevancy and diversity of the returned list of ranked images.  ...  This paper summarizes the participation of Imcube at the Retrieving Diverse Social Images Task of MediaEval 2015.  ...  Visual relevancy improvement Visual information is also used to improve the relevancy by reranking them according to different criteria.  ... 
dblp:conf/mediaeval/SchmiedekeKG15 fatcat:dvw54qu5ffgqpmbfijdkmhgsmq

Re-ranking with Click-Base Similarity and Typicality using Spectral Clustering

Mayuri Kawalkar, Gangotri Nathaney
2017 IARJSET  
A hyper graph is used to model the relationship between images by integrating low level feature and semantic feature. For visually similar images the feature are extracted.  ...  Saliency region contain more important information in the image. So less time require for searching.  ...  To improve image categorization result proposed Re-Ranking based image searching using saliencing technique. Re-ranking provide proper image classification.  ... 
doi:10.17148/iarjset.2017.4611 fatcat:cqvqbdkumzhptjfpqwbz2diwsi

A Pose-Sensitive Embedding for Person Re-Identification with Expanded Cross Neighborhood Re-Ranking [article]

M. Saquib Sarfraz, Arne Schumann, Andreas Eberle, Rainer Stiefelhagen
2018 arXiv   pre-print
well by using simple direct rank list based comparison or even by just using the already computed euclidean distances between the images.  ...  We show that in contrast to the current state-of-the-art re-ranking methods our approach does not require to compute new rank lists for each image pair (e.g., based on reciprocal neighbors) and performs  ...  Acknowledgement: The research described in this work is funded in part by the BMBF grant No. 13N14029.  ... 
arXiv:1711.10378v2 fatcat:mkfpqpbonbg7tkzkc42fvxsmcq

A Pose-Sensitive Embedding for Person Re-identification with Expanded Cross Neighborhood Re-ranking

M. Saquib Sarfraz, Arne Schumann, Andreas Eberle, Rainer Stiefelhagen
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
well by using simple direct rank list based comparison or even by just using the already computed euclidean distances between the images.  ...  We show that in contrast to the current state-of-the-art re-ranking methods our approach does not require to compute new rank lists for each image pair (e.g., based on reciprocal neighbors) and performs  ...  Acknowledgement: The research described in this work is funded in part by the BMBF grant No. 13N14029.  ... 
doi:10.1109/cvpr.2018.00051 dblp:conf/cvpr/SarfrazSES18 fatcat:flwpklkpyvakpoos4zwnk4f6f4

Semantic Smoothing of Document Models for Agglomerative Clustering

Xiaohua Zhou, Xiaodan Zhang, Xiaohua Hu
2007 International Joint Conference on Artificial Intelligence  
In this framework, candidate images are retrieved at first by considering their associated text information.  ...  This paper proposes a new framework for Web search, which exploits image contents to help improve the search performance.  ...  This work was supported by the National Science Foundation of China ( 60505013 ), the National Science Fund for Distinguished Young Scholars of China (60325207), the Jiangsu Science Foundation (BK2005412  ... 
dblp:conf/ijcai/ZhouZH07 fatcat:2ku4b5lcpbclrhvydq33tvkmzq

Combining re-ranking and rank aggregation methods for image retrieval

Daniel Carlos Guimarães Pedronette, Ricardo da S. Torres
2015 Multimedia tools and applications  
Given a query image as input, CBIR systems retrieve the most similar images in a collection by taking into account image visual properties.  ...  This paper presents novel approaches for combining re-ranking and rank aggregation methods aiming at improving the effectiveness of Content-Based Image Retrieval (CBIR) systems.  ...  These approaches propose improving the effectiveness of image searches by exploiting the information about the relationships among images, commonly referred to as contextual information, in an unsupervised  ... 
doi:10.1007/s11042-015-3044-0 fatcat:pvqdrgor5jfpzn2y7tdfcwsp2q

Ranking user-annotated images for multiple query terms

Moray Allan, Jakob Verbeek
2009 Procedings of the British Machine Vision Conference 2009  
Search queries are answered by using a mixture of kernel density estimators to rank the visual content of web images from the Flickr website whose noisy tag annotations match the given query terms.  ...  We show how web image search can be improved by taking into account the users who provided different images, and that performance when searching for multiple terms can be increased by learning a new combined  ...  Acknowledgements: this work has been partially funded by the ANR project R2I. References  ... 
doi:10.5244/c.23.20 dblp:conf/bmvc/AllanV09 fatcat:jnkek7iw7razbi2iw4jlqic7dy

Recod @ MediaEval 2016: Diverse Social Images Retrieval

Cristiano D. Ferreira, Rodrigo Tripodi Calumby, Iago B. A. do C. Araujo, Ícaro C. Dourado, Javier A. V. Muñoz, Otávio Augusto Bizetto Penatti, Lin Tzy Li, Jurandy Almeida, Ricardo da Silva Torres
2016 MediaEval Benchmarking Initiative for Multimedia Evaluation  
Our proposal is based on re-ranking, rank aggregation, and diversity promotion, allowing employment of textual and visual information apart or fused.  ...  This paper presents the RECOD team experience in the Retrieving Diverse Social Images Task at MediaEval 2016. The teams were required to develop a diversification approach for social photo retrieval.  ...  These improved ranked lists were used as input for a clustering-based summarization method.  ... 
dblp:conf/mediaeval/FerreiraCADMPLA16 fatcat:hh4cuzdo7faolgjigr2vgjnqyq

Survey on Methodologies Used for Web Image Search

Dipalee N.Gaikwad, Sandeep U. Kadam
2014 International Journal of Computer Applications  
Also some methods are depend on the visual similarities between the images. To improve the result of web image search, strategies like keyword expansion, active re-ranking is also used.  ...  In this paper the methods developed by different researchers in the area of web image search are reviewed. These methods vary from textual information search to user feedback.  ...  The future work for image search is improved in the area of re-ranking using duplication detection. Using the combination of textual and visual features search is explored.  ... 
doi:10.5120/18853-0416 fatcat:yf6qqpdadbf6rbozqfhyyk4k4e

Rescue Tail Queries

Xiaopeng Yang, Tao Mei, Yongdong Zhang
2014 Proceedings of the ACM International Conference on Multimedia - MM '14  
Image search engines have achieved good performance for head (popular) queries by leveraging text information and user click data.  ...  To deal with this challenge, we propose to mine relevant information from the very few click data by leveraging click-wise-based image pairs and query-dependent multimodal fusion.  ...  This further shows the superiority of relevant information mined from click-wise-based image pairs used in our approach.  ... 
doi:10.1145/2647868.2654900 dblp:conf/mm/YangMZ14 fatcat:plpat2pfrjgqtbqwiad4whqowi

GeoTime-based tag ranking model for automatic image annotation

Mădălina Mitran, Guillaume Cabanac, Mohand Boughanem
2014 Proceedings of the 29th Annual ACM Symposium on Applied Computing - SAC '14  
The experimental results showed that our GeoTimebased tag ranking model yields significant improvement over two state-of-the-art baselines.  ...  Finally, the top-ranked tags are used to annotate a query image. We conducted a series of experiments on a dataset consisting of over 201,000 Flickr images from the Paris geographic area.  ...  In our work we used data from image sharing platforms and we focused on textual information provided by these ones and on temporal and spatial information stored in the image EXIF descriptor.  ... 
doi:10.1145/2554850.2554866 dblp:conf/sac/MitranCB14 fatcat:s4jbkh6trrftdhvdpmndkp4jhu

Tag Quality Improvement for Social Image Hosting Website

Jiyi Li, Qiang Ma, Yasuhito Asano, Masatoshi Yoshikawa
2013 IPSJ Online Transactions  
Users can upload and tag their images or search for images by using keywords which describe image semantics.  ...  To improve tag quality, we propose four approaches with one framework to automatically generate new tags, and rank the new tags as well as the existing raw tags, for both untagged and tagged images.  ...  [5] proposed by Liu et al. is related to the case of unranked tag. It ranks the raw tags of images by using a linear combination on textual and visual information with a random walk process.  ... 
doi:10.2197/ipsjtrans.6.75 fatcat:goiyplvmjvh55gurnmjchcub4m

Exploiting contextual information for image re-ranking and rank aggregation

Daniel Carlos Guimarães Pedronette, Ricardo da S. Torres
2012 International Journal of Multimedia Information Retrieval  
This paper presents a novel re-ranking algorithm aiming to exploit contextual information for improving the effectiveness of rankings computed by CBIR systems.  ...  In our approach, ranked lists and distance scores are used to create context images, later used for retrieving contextual information.  ...  We aim to characterize contextual information by analyzing this region using image processing techniques. These information will be used by the re-ranking method presented in next section.  ... 
doi:10.1007/s13735-012-0002-8 fatcat:uoicbakbyfe2vb72rtrlqudf3m

Improving Web Image Search Re-ranking

L Ramadevi
2013 IOSR Journal of Computer Engineering  
The Extensive experiments demonstrate that using this algorithm to filter output of Google Image Search and Microsoft Live Image Search is a practical and effective way to dramatically improve the user's  ...  A query image is first categorized as one of several predefined intention categories, and a specific similarity measure has been used inside each of category to combine the image features for re-ranking  ...  Since both image understanding and local text information are exploited by the image search engine, we wonder if we can re-rank the image list using global information, i.e. text in the HTML document,  ... 
doi:10.9790/0661-1314450 fatcat:cux3f3smy5bw3paiiayv3k4aje
« Previous Showing results 1 — 15 out of 623,469 results