[Bài báo] Logo Matching for Document Image Retrieval

... one definition for each possible word sense. The algo- rithm looks for words in the sense definitions that overlap with context words in the given sentence, and chooses the sense that yields maximal word overlap. ... of the 25 words in the Senseval sample as a target word for the sense matching task. Next, we had to pick for each target word a corresponding synonym to p...

Ngày tải lên: 20/02/2014, 12:20

... starting point for indexing. We have focused on the devel- opment and implementation of automatic caption- based techniques for indexing and retrieval of pho- tographs taken at scenes of crime (SOC). Researchers ... NLP for Indexing and Retrieval of Captioned Photographs Katerina Pastra, Horacio Saggion, Yorick Wilks Department of Computer Science University...

Ngày tải lên: 22/02/2014, 02:20

... the ACL-IJCNLP 2009 Conference Short Papers, pages 193–196, Suntec, Singapore, 4 August 2009. c 2009 ACL and AFNLP A Rose is a Roos is a Ruusu: Querying Translations for Web Image Search Janara ... 22% better images. 2 Background IDIOM makes heavy use of a sense disambiguated, vastly multilingual dictionary called PANDIC- TIONARY (Mausam et al., 2009). PANDIC- TIO...

Ngày tải lên: 08/03/2014, 01:20

... score for summary quality was not. For the Mitre data, the differences in score between the Concat and Sent systems for grammati- cality and summary quality were statistically signif- icant at the ... text to assist aphasic readers. In Proceedings of the AAAI-98 Workshop on Integrating Artificial Intelligence and Assistive Technology. R. Chandrasekar, Christy Doran, and Srinivas Bangalor...

Ngày tải lên: 08/03/2014, 07:20

... our approach against a baseline that compares can- didate sentences not against patterns that were gained from question/answer sentence pairs, but from questions alone. In order to create these pat- terns, ... answer sentences. In this paper we describe an algorithm that learns possible syntactic answer sentence formu- lations for syntactic question classes from a set of example...

Ngày tải lên: 17/03/2014, 22:20

... natural language processing for the tasks of document sum- marization (Lin and Bilmes, 2010) and word align- ment (Lin and Bilmes, 2011). Submodular functions share a number of proper- ties in common ... convolutions). For exam- ple, if a collection of functions {F i } i is submodular, then so is their weighted sum F =  i α i F i where α i are nonnegative weights. It is n...

Ngày tải lên: 23/03/2014, 16:20

... the best of the 4 models for entity annotation which have been evaluated. 8 Related Research Using text that accompanies the image for annotat- ing images and for training image recognition is not new. ... 2007. c 2007 Association for Computational Linguistics Text Analysis for Automatic Image Annotation Koen Deschacht and Marie-Francine Moens Interdisciplinary Centre fo...

Ngày tải lên: 23/03/2014, 18:20

... Republic, June 2007. c 2007 Association for Computational Linguistics User Requirements Analysis for Meeting Information Retrieval Based on Query Elicitation Vincenzo Pallotta Department of ... implications of these findings for the design of meeting retrieval systems. Section 4 concludes the paper and outlines some directions for future work. 2 Argumentative...

Ngày tải lên: 23/03/2014, 18:20

... Toronto, Ontario, Canada, April 1991. [10] A. A. Yeni and S. Ert ă urk, Fast digital image stabilization using one bit transform based sub -image motion estimation, ” IEEE Transactions on Consumer ... Global Motion Estimation Based on Laplacian Two-Bit Plane Matching for Fast Digital Image Stabilization Nam-Joon Kim, 1 Hyuk-Jae Lee, 1 and Jae-Beom Lee 2...

Ngày tải lên: 22/06/2014, 00:20

... com- bining the output of five different expert classifiers trained using three different low-level features. In the system intro- duced in [12], several low-level image primitives were com- bined in ... importance for each feature in the under- lying linear combination. The work described in this paper focuses on obtaining an “optimal” metrics based on a lin- ear combination o...

Ngày tải lên: 22/06/2014, 06:20

... Toronto, Ontario, Canada, April 1991. [10] A. A. Yeni and S. Ert ă urk, Fast digital image stabilization using one bit transform based sub -image motion estimation, ” IEEE Transactions on Consumer ... Energy. REFERENCES [1] S. Ert ă urk, Digital image stabilization with sub -image phase correlation based global motion estimation, ” IEEE Transac- tions on Consumer E...

Ngày tải lên: 22/06/2014, 06:20

... to estimate what the user actually wants. Many types of learn- ing models have been applied in relevance feedback for image retrieval, such as Bayesian framework [6–8], SVM [9], and active learning ... BASED ON A STATE-SPACE MODEL 4.1. State-space model The state-space approach has been widely applied to the analysis of dynamic systems, which involve estimating the state of...

Ngày tải lên: 22/06/2014, 19:20

... Journal on Advances in Signal Processing Volume 2007, Article ID 43450, 17 pages doi:10.1155/2007/43450 Research Article An Attention-Driven Model for Grouping Similar Images with Image Retrieval ... specific ideas about the images they are searching for. Clustering techniques can be used to or- ganize visually similar images into groups and mini- mize the numb...

Ngày tải lên: 22/06/2014, 23:20

... registration, shape matching and shape recognition. In our experiments, we have demonstrated BELONGIE ET AL.: SHAPE MATCHING AND OBJECT RECOGNITION USING SHAPE CONTEXTS 519 Fig. 11. Examples of shapes ... asgivenin(19)comparedtoa BELONGIE ET AL.: SHAPE MATCHING AND OBJECT RECOGNITION USING SHAPE CONTEXTS 517 4. DeCoste and Scho È lkopf [13] report an...

Ngày tải lên: 31/12/2014, 05:35

... underlying transformations (affine for linear and TPS for non-linear transformation), which improves shape matching and discrimination. Shape contexts method gives the best logo matching per- formance ... detected logos, our goal of logo matching is to compute an effective ranked list for logos in the database. By constructing the list of best matching logos, we effectively r...

Ngày tải lên: 31/12/2014, 06:07