206,623 Hits in 3.6 sec

Automatic image annotation using visual content and folksonomies

Stefanie Lindstaedt, Roland Mörzinger, Robert Sorschag, Viktoria Pammer, Georg Thallinger
2008 Multimedia tools and applications  
Our approach applies two techniques based on image analysis: Classification annotates images with a controlled vocabulary while tag propagation uses user generated, folksonomic annotations and is therefore  ...  This paper describes techniques for automatic image annotation by taking advantage of collaboratively annotated image databases, so called visual folksonomies.  ...  In this setup we intentionally included visually rather distinct concepts as well as visually similar concepts like raspberry and strawberry.  ... 
doi:10.1007/s11042-008-0247-7 fatcat:5z3earvzyjd2va66l64r4pg7ai

Score Propagation Based on Similarity Shot Graph for Improving Visual Object Retrieval

Juan Manuel Barrios, Jose Manuel Saavedra
2015 Proceedings of the Third Edition Workshop on Speech, Language & Audio in Multimedia - SLAM '15  
Four methods for creating the SSG are presented: two based on computing and comparing low-level visual features, one based on comparing text transcriptions, and other based on computing and comparing high-level  ...  concepts.  ...  The graph connects shots according to some criterion, like low-level visual similarity, similarity of speech or subtitles, or even similarity based on shared semantic concepts.  ... 
doi:10.1145/2802558.2814644 dblp:conf/mm/BarriosS15 fatcat:bgggwovc3nexfah5reqvwijjqu

A Visual Annotation Framework Using Common-Sensical and Linguistic Relationships for Semantic Media Retrieval [chapter]

Bageshree Shevade, Hari Sundaram
2006 Lecture Notes in Computer Science  
We also develop a new evaluation technique for annotation that is based on relationship between concepts based on commonsensical relationships.  ...  In this paper, we present a novel image annotation approach with an emphasis on -(a) common sense based semantic propagation, (b) visual annotation interfaces and (c) novel evaluation schemes.  ...  The user can also add text annotations to these visual concepts. The system then propagates these annotations to all the visual concept clusters based on low-level features and WordNet.  ... 
doi:10.1007/11670834_20 fatcat:gkjp25tgajgopkvfhdrisp5wam

Video search re-ranking via multi-graph propagation

Jingjing Liu, Wei Lai, Xian-Sheng Hua, Yalou Huang, Shipeng Li
2007 Proceedings of the 15th international conference on Multimedia - MULTIMEDIA '07  
To better exploit the underlying relationship between video shots, the proposed reranking scheme simultaneously leverages textual relevancy, semantic concept relevancy, and low-level-feature-based visual  ...  A modified topic-sensitive PageRank algorithm is then applied on these graphs to propagate the relevance scores through all related video shots.  ...  Fig. 3 . 3 A graph based on a specific concept "car." Fig. 4 . 4 A graph pruned based on visual similarity. Fig. 5 . 5 A graph re-constructed with directed hyperlinks. Fig.  ... 
doi:10.1145/1291233.1291279 dblp:conf/mm/LiuLHHL07 fatcat:sp47lkrnmzegzpcfxootw7whxm

Interactive User Feedback in Ontology Matching Using Signature Vectors

Isabel F. Cruz, Cosmin Stroe, Matteo Palmonari
2012 2012 IEEE 28th International Conference on Data Engineering  
This feedback process has been implemented in the AgreementMaker system and is supported by visual analytic techniques that help users to better understand the matching process.  ...  Feedback Propagation. This component is responsible for the propagation of the user feedback on one mapping to other mappings using the Similarity Update and Propagation method.  ...  This process uses the Signature-based Mapping Clustering and Similarity Update and Propagation methods. Disagreement-based Top-k Mapping Selection.  ... 
doi:10.1109/icde.2012.137 dblp:conf/icde/CruzSP12 fatcat:zzkfitdsq5bszeuafsggh3hs5i

Towards Training-Free Refinement for Semantic Indexing of Visual Media [chapter]

Peng Wang, Lifeng Sun, Shiqang Yang, Alan F. Smeaton
2016 Lecture Notes in Computer Science  
Indexing of visual media based on content analysis has now moved beyond using individual concept detectors and there is now a focus on combining concepts or post-processing the outputs of individual concept  ...  In contrast to training-dependent methods which dominate this field, this paper presents a training-free refinement (TFR) algorithm for enhancing semantic indexing of visual media based purely on concept  ...  After a number of iterations, the algorithm converges to a solution in which the last row of C n is a prediction based on similarity propagation.  ... 
doi:10.1007/978-3-319-27671-7_21 fatcat:455uhtskanbbjouz6w2227zzam

Multi-modality web video categorization

Linjun Yang, Jiemin Liu, Xiaokang Yang, Xian-Sheng Hua
2007 Proceedings of the international workshop on Workshop on multimedia information retrieval - MIR '07  
The semantic modality includes three feature representations, i.e., concept histogram, visual word vector model and visual word Latent Semantic Analysis (LSA), while text modality includes the titles,  ...  (2) SVM outperforms GMM and MR on nearly all the modalities.  ...  The other is based on visual words (similar to visual terms in [21] ) which can be regarded as implicit concepts. Visual words are the cluster labels of shots achieved by clustering.  ... 
doi:10.1145/1290082.1290119 dblp:conf/mir/YangLYH07 fatcat:lpiiw2e3kvbuleejndhhcns3zu

Training-free indexing refinement for visual media via multi-semantics

Peng Wang, Lifeng Sun, Shiqiang Yang, Alan F. Smeaton
2017 Neurocomputing  
* factorization and neighbourhood-based graph propagation are both used in the refinement of semantics.  ...  Furthermore, any available ontological concept relationships among concepts can also be integrated into this model as an additional source of external a priori knowledge.  ...  Indexing of visual media based on content analysis has now moved beyond using individual concept detectors and there is now a focus on combining concepts by post-processing the outputs of individual concept  ... 
doi:10.1016/j.neucom.2016.08.107 fatcat:u33uvlfc2bclrkyckr4c6mqmu4

On the pooling of positive examples with ontology for visual concept learning

Shiai Zhu, Chong-Wah Ngo, Yu-Gang Jiang
2011 Proceedings of the 19th ACM international conference on Multimedia - MM '11  
A common obstacle in effective learning of visual concept classifiers is the scarcity of positive training examples due to expensive labeling cost.  ...  This effectively widens the coverage of the positive samples with visually more diversified content, which is important for learning a good concept classifier.  ...  The second measure is based on ontological similarity.  ... 
doi:10.1145/2072298.2071934 dblp:conf/mm/ZhuNJ11 fatcat:h4w2svg5onanboqzmyk4hwuhua

Annotating personal albums via web mining

Jimin Jia, Nenghai Yu, Xian-Sheng Hua
2008 Proceeding of the 16th ACM international conference on Multimedia - MM '08  
A multi-graph similarity propagation based semisupervised learning (MGSP-SSL) algorithm is used to suppress the noises in the initial annotations from the Web.  ...  Existing research on image annotation evolves through two stages: learning-based methods and web-based methods.  ...  For an image of any kind, it is difficult to provide abundant and accurate annotations based on such limited-size concept sets.  ... 
doi:10.1145/1459359.1459421 dblp:conf/mm/JiaYH08 fatcat:rbmqunfgrjbevdgfvypkt3fclm

Spatiotemporal semantic video segmentation

E. Galmar, Th. Athanasiadis, B. Huet, Y. Avrithis
2008 2008 IEEE 10th Workshop on Multimedia Signal Processing  
Then, we iteratively merge consecutive blocks by a matching procedure which considers both semantic and visual properties. Results on real video sequences show the potential of our approach.  ...  An internal graph structure that describes both visual and semantic properties of image and video regions is adopted.  ...  The criterion to match a dominant volume a to a volume b, e ab ∈ E a , is based on both semantic and visual attributes. Let c * a and c * b be the dominant concepts of L a and L b .  ... 
doi:10.1109/mmsp.2008.4665143 dblp:conf/mmsp/GalmarAHA08 fatcat:yx53wrwp7bas7pw3ichhshvztm

A Scalable Graph-Based Semi-Supervised Ranking System for Content-Based Image Retrieval

Xiaojun Qi, Ran Chang
2013 International Journal of Multimedia Data Engineering and Management  
Two-layer manifold graphs are then built in both low-level visual and high-level semantic spaces. One graph is constructed at the first layer using anchor images obtained from the feedback log.  ...  These vectors are fused to propagate relevance scores of labeled images to unlabeled images.  ...  and second layer graphs; Variant 5: The CBIR system that incorporates L 1 -based low-level visual and high-level semantic similarities into the first layer graph and L 1 -based low-level visual similarity  ... 
doi:10.4018/ijmdem.2013100102 fatcat:c4b76naaejhyhcwxghskmk7lte

CSRNCVA: A model of cross-media semantic retrieval based on neural computing of visual and auditory sensations

Yang Liu, Kun Cai, Chun Liu, Fengbin Zheng
2018 Neural Network World  
Algorithms based on CSRNCVA were developed. It employs belief propagation algorithms of probabilistic graphical model and hierarchical learning.  ...  Considering an idea from deep belief network and hierarchical temporal memory, we presented a brain-inspired intelligent model, called cross-media semantic retrieval based on neural computing of visual  ...  similar concept of visual-auditory integration CTS which meets the condition that CM ≈ 0 for the concept of visual-auditory integration CTS.  ... 
doi:10.14311/nnw.2018.28.018 fatcat:e6kg3h5f2zaqpbotjn2aty5y4e

Integrating Concept Ontology and Multitask Learning to Achieve More Effective Classifier Training for Multilevel Image Annotation

Jianping Fan, Yuli Gao, Hangzai Luo
2008 IEEE Transactions on Image Processing  
Index Terms-Concept ontology, hierarchical boosting, interactive hypotheses assessment, interconcept visual similarity, intraconcept visual diversity, multiple kernel learning, multitask learning.  ...  To tackle the problem of huge intraconcept visual diversity, multiple types of kernels are integrated to characterize the diverse visual similarity relationships between the images more precisely, and  ...  As shown in Fig. 11 , one can observe that good mixture-of-kernels can make the visually-similar images to be displayed closely according to their kernel-based visual similarity, and the visually-dissimilar  ... 
doi:10.1109/tip.2008.916999 pmid:18270128 fatcat:bxokhjalpzcfdpwyk3yajzl3ni

Comparison of clustering approaches for summarizing large populations of images

Yushi Jing, Michele Covell, Henry A. Rowley
2010 2010 IEEE International Conference on Multimedia and Expo  
We improve on these early results using a simple distribution-based selection filter on incomplete clustering results.  ...  This paper compares the efficacy and efficiency of different clustering approaches for selecting a set of exemplar images, to present in the context of a semantic concept.  ...  CONCLUSIONS AND FUTURE WORK Our results suggest that, at least for a subset of concepts that have a strong visual component, cluster-based selection of exemplars can get surprisingly close to the same  ... 
doi:10.1109/icme.2010.5583276 dblp:conf/icmcs/JingCR10 fatcat:h5tbugxmlbfjxkon7xisqn75jm
« Previous Showing results 1 — 15 out of 206,623 results