18,502 Hits in 4.8 sec

Exploiting visual word co-occurrence for image retrieval

Miaojing Shi, Xinghai Sun, Dacheng Tao, Chao Xu
2012 Proceedings of the 20th ACM international conference on Multimedia - MM '12  
A co-occurrence matrix is introduced to refine the similarity measure for image ranking.  ...  In this paper, we construct a visual word co-occurrence table by exploring visual word co-occurrence extracted from small affine-invariant regions in a large collection of natural images.  ...  CONCLUSIONS This paper has proposed a novel image retrieval approach that exploits the spatial co-occurrence of visual words.  ... 
doi:10.1145/2393347.2393364 dblp:conf/mm/ShiSTX12 fatcat:kcv53roeqbahzljcbnaudhjrsq

Exploring Spatial Correlation for Visual Object Retrieval

Miaojing Shi, Xinghai Sun, Dacheng Tao, Chao Xu, George Baciu, Hong Liu
2015 ACM Transactions on Intelligent Systems and Technology  
Subsequently, we propose two methods of co-occurrence weighting similarity measure for image ranking: Co-Cosine and Co-TFIDF.  ...  This paper outlines the construction of a visual word co-occurrence matrix by exploring visual word co-occurrence extracted from small affine-invariant regions in a large collection of natural images.  ...  In comparison, our method only exploits the spatial co-occurrence of visual words in the entire dataset.  ... 
doi:10.1145/2641576 fatcat:jibn2tavdjezhbrr6wzdvy2x6e

Toward a higher-level visual representation for object-based image retrieval

Yan-Tao Zheng, Shi-Yong Neo, Tat-Seng Chua, Qi Tian
2008 The Visual Computer  
We propose a higher-level visual representation, visual synset, for object-based image retrieval beyond visual appearances.  ...  First, the approach strengthens the discrimination power of visual words by constructing an intermediate descriptor, visual phrase, from frequently co-occurring visual word-set.  ...  Phase 2 tackles the low discrimination issue in visual words, by exploiting the spatial co-occurrence information among visual words.  ... 
doi:10.1007/s00371-008-0294-0 fatcat:mtmoj5gysrb43e32jffbcvdsfy

Integrating visual and semantic contexts for topic network generation and word sense disambiguation

Jianping Fan, Hangzai Luo, Yi Shen, Chunlei Yang
2009 Proceeding of the ACM International Conference on Image and Video Retrieval - CIVR '09  
for addressing the issues of polysemes and synonyms more effectively, thus it can significantly improve the precision and recall rates for image retrieval.  ...  Second, our word sense disambiguation algorithm can integrate the topic network to exploit both the visual similarity contexts between the images and the semantic similarity contexts between their tags  ...  co-occurrence probability.  ... 
doi:10.1145/1646396.1646440 dblp:conf/civr/FanLSY09 fatcat:dvczr67ddnenjhp3w3reurc25u

Improving Content-Based Image Retrieval by Identifying Least and Most Correlated Visual Words [chapter]

Leszek Kaliciak, Dawei Song, Nirmalie Wiratunga, Jeff Pan
2012 Lecture Notes in Computer Science  
In this paper, we propose a model for direct incorporation of image content into a (short-term) user profile based on correlations between visual words and adaptation of the similarity measure.  ...  The information about the most and the least correlated visual words can be exploited in order to adapt the similarity measure.  ...  Liu et. al. ([6] ) exploit co-occurrence information in spatial domain. Authors make an assumption that the related visual words would appear in a certain neighbourhood.  ... 
doi:10.1007/978-3-642-35341-3_27 fatcat:m5tfbwy6rbajhnx6jc5xx4he5y

Consensus-Aware Visual-Semantic Embedding for Image-Text Matching [article]

Haoran Wang, Ying Zhang, Zhong Ji, Yanwei Pang, Lin Ma
2021 arXiv   pre-print
Specifically, the consensus information is exploited by computing the statistical co-occurrence correlations between the semantic concepts from the image captioning corpus and deploying the constructed  ...  Afterwards, CVSE learns the associations and alignments between image and text based on the exploited consensus as well as the instance-level representations for both modalities.  ...  With the instantiated concepts, their co-occurrence relationship are examined to build one correlation graph and thereby exploit the commonsense knowledge.  ... 
arXiv:2007.08883v2 fatcat:xw4ktscovvdz3fuhe3vkjstbie

Combining Multi-modal Features for Social Media Analysis [chapter]

Spiros Nikolopoulos, Eirini Giannakidou, Ioannis Kompatsiaris, Ioannis Patras, Athena Vakali
2011 Social Media Modeling and Computing  
Tagged images taken from social sites have been used in the characteristic scenarios of image clustering and retrieval, to demonstrate the benefits of multi-modal analysis in social media.  ...  In this chapter we discuss methods for efficiently modeling the diverse information carried by social media.  ...  For all images the authors released 500-dimensional co-occurrence vectors for visual words (as described in Section 4.2.1), as well as 1000-dimensional co-occurrence vectors for tag-words (as described  ... 
doi:10.1007/978-0-85729-436-4_4 fatcat:c5rsc2fi5zcglcmgtvsa5b6fca

Multimodal Distributional Semantics

E. Bruni, N. K. Tran, M. Baroni
2014 The Journal of Artificial Intelligence Research  
of a word can be extended to also encompass its co-occurrence with the visual words of images it is associated with.  ...  Distributional semantic models derive computational representations of word meaning from the patterns of co-occurrence of words in text.  ...  In particular, our multimodal distributional semantic model (MDSM) exploits both co-occurrence with words (from a standard text corpus) and co-occurrence with visual features extracted using computer vision  ... 
doi:10.1613/jair.4135 fatcat:hwiocijrfbfdbhufc26eyaxgim

Automatic Image Annotation Using Maximum Entropy Model [chapter]

Wei Li, Maosong Sun
2005 Lecture Notes in Computer Science  
Automatic image annotation is a newly developed and promising technique to provide semantic image retrieval via text descriptions.  ...  It concerns a process of automatically labeling the image contents with a pre-defined set of keywords which are exploited to represent the image semantics.  ...  By exploiting text and image feature co-occurrence statistics, these methods can extract hidden semantics from images, and have been proven successful in constructing a nice framework for the domain of  ... 
doi:10.1007/11562214_4 fatcat:4mkfcthddbcdfchkvsxynczywi

A Bayesian Approach to Hybrid Image Retrieval [chapter]

Pradhee Tandon, C. V. Jawahar
2009 Lecture Notes in Computer Science  
Content based image retrieval (CBIR) has been well studied in the computer vision and multimedia community.  ...  We fuse both of them in a Bayesian framework to design a hybrid image retrieval system by overcoming their shortcomings.  ...  On contrary, CFIR methods exploit the co-occurrence information (for example in a collaborative filtering framework) in the logs of image-access to model the similarity across images [3, 5] .  ... 
doi:10.1007/978-3-642-11164-8_79 fatcat:ja6c64prxzgzxoip52xykgooji

High order pLSA for indexing tagged images

S. Nikolopoulos, S. Zafeiriou, I. Patras, I. Kompatsiaris
2013 Signal Processing  
This work presents a method for the efficient indexing of tagged images. Tagged images are a common resource of social networks and occupy a large portion of the social media stream.  ...  Their basic characteristic is the co-existence of two heterogeneous information modalities, i.e. visual and tag, which refer to the same abstract meaning.  ...  For all images we have used the 500-dimensional co-occurrence vectors for visual words and the 1000-dimensional cooccurrence vectors for tag-words that were released by the authors.  ... 
doi:10.1016/j.sigpro.2012.08.004 fatcat:2bacd2atq5eivn6ths6ehftzc4

Semantic-visual concept relatedness and co-occurrences for image retrieval

Linan Feng, Bir Bhanu
2012 2012 19th IEEE International Conference on Image Processing  
Third, we leverage the visual and semantic correspondence and the co-occurrence patterns to improve the accuracy and efficiency for image retrieval.  ...  This paper introduces a novel approach that allows the retrieval of complex images by integrating visual and semantic concepts. The basic idea consists of three aspects.  ...  Therefore, the measure from concept co-occurrence could be more important. The main contributions of this paper is to uncover concept co-occurrence patterns for image retrieval.  ... 
doi:10.1109/icip.2012.6467388 dblp:conf/icip/FengB12 fatcat:e3su7ccs3vdljnkcq4ytulv5de

Negative Evidences and Co-occurences in Image Retrieval: The Benefit of PCA and Whitening [chapter]

Hervé Jégou, Ondřej Chum
2012 Lecture Notes in Computer Science  
and the co-occurrences phenomenon.  ...  We show and explicitly exploit relations between i) mean subtraction and the negative evidence, i.e., a visual word that is mutually missing in two descriptions being compared, and ii) the axis de-correlation  ...  The paper is organized as follows: After introducing the context in Section 2, Section 3 shows the role of co-missing visual words and Section 4 exploits whitening to address the issue arising with co-occurrence  ... 
doi:10.1007/978-3-642-33709-3_55 fatcat:xqajeeyjo5f6td4pysnz3zlun4

Multimodal pLSA on visual features and tags

Stefan Romberg, Eva Horster, Rainer Lienhart
2009 2009 IEEE International Conference on Multimedia and Expo  
This work studies a new approach for image retrieval on largescale community databases.  ...  We use topic models to derive a high-level representation appropriate for retrieval for each of our images in the database.  ...  The occurrences of visual words in an image are hereby counted into a co-occurrence vector, also called document vector.  ... 
doi:10.1109/icme.2009.5202522 dblp:conf/icmcs/RombergHL09 fatcat:kfremgkw3bgq3kn3mriwutqa3q

Quadtree decomposition based extended vector space model for image retrieval

Vignesh Ramanathan, Shaunak Mishra, Pabitra Mitra
2011 2011 IEEE Workshop on Applications of Computer Vision (WACV)  
Bag of visual words approach for image retrieval does not exploit the spatial distribution of visual words in an image.  ...  This paper proposes a novel extended vector space based image retrieval technique which takes into account the spatial occurrence (context) of a visual word in an image along with the co-occurrence of  ...  Hence, comparative results with SPM based scheme are provided to evaluate our extended vector space approach for image retrieval based on spatial distribution and co-occurrences of visual words.  ... 
doi:10.1109/wacv.2011.5711495 dblp:conf/wacv/RamanathanMM11 fatcat:dojp7bnwc5dfzaqzbw6j4pbm2m
« Previous Showing results 1 — 15 out of 18,502 results