2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR)最新文献_第2页

Segmentation-Free Speech Text Recognition for Comic Books 漫画书的无分割语音文本识别

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) Pub Date : 2017-11-09 DOI: 10.1109/ICDAR.2017.288

Christophe Rigaud, J. Burie, J. Ogier

引用次数: 15

Preparatory KWS Experiments for Large-Scale Indexing of a Vast Medieval Manuscript Collection in the HIMANIS Project 在HIMANIS计划中大规模索引大量中世纪手稿集的预备KWS实验

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) Pub Date : 2017-11-09 DOI: 10.1109/ICDAR.2017.59

Théodore Bluche, Sébastien Hamel, Christopher Kermorvant, J. Puigcerver, D. Stutzmann, A. Toselli, E. Vidal

{"title":"Preparatory KWS Experiments for Large-Scale Indexing of a Vast Medieval Manuscript Collection in the HIMANIS Project","authors":"Théodore Bluche, Sébastien Hamel, Christopher Kermorvant, J. Puigcerver, D. Stutzmann, A. Toselli, E. Vidal","doi":"10.1109/ICDAR.2017.59","DOIUrl":"https://doi.org/10.1109/ICDAR.2017.59","url":null,"abstract":"Making large-scale collections of digitized historical documents searchable is being earnestly demanded by many archives and libraries. Probabilistically indexing the text images of these collections by means of keyword spotting techniques is currently seen as perhaps the only feasible approach to meet this demand. A vast medieval manuscript collection, written in both Latin and French, called \"Chancery\", is currently being considered for indexing at large. In addition to its bilingual nature, one of the major difficulties of this collection is the very high rate of abbreviated words which, on the other hand, are completely expanded in the ground truth transcripts available. In preparation to undertake full indexing of Chancery, experiments have been carried out on a relatively small but fully representative subset of this collection. To this end, a keyword spotting approach has been adopted which computes word relevance probabilities using character lattices produced by a recurrent neural network and a N-gram character language model. Results confirm the viability of the chosen approach for the large-scale indexing aimed at and show the ability of the proposed modeling and training approaches to properly deal with the abbreviation difficulties mentioned.","PeriodicalId":433676,"journal":{"name":"2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128033051","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 38

A Spatial Domain Steganography for Grayscale Documents Using Pattern Recognition Techniques 基于模式识别技术的灰度文档空间域隐写

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) Pub Date : 2017-11-09 DOI: 10.1109/ICDAR.2017.391

J. Burie, J. Ogier, Cu Vinh Loc

{"title":"A Spatial Domain Steganography for Grayscale Documents Using Pattern Recognition Techniques","authors":"J. Burie, J. Ogier, Cu Vinh Loc","doi":"10.1109/ICDAR.2017.391","DOIUrl":"https://doi.org/10.1109/ICDAR.2017.391","url":null,"abstract":"Steganography is an effective way to hide a secret message into a document image with the objective of providing authenticity of transmitted documents. Steganography has been widely used for natural images but few researches have been carried out to apply this strategy on document images. In this study, we proposed a novel data hiding scheme that enables to embed a secret information with moderate length by taking advantages of pattern recognition techniques. Firstly, the potential feature points used for constructing embedding regions are identified by using the Speed Up Robust Features (SURF) detector. Secondly, Local Binary Pattern (LBP) is utilized to figure out embedding patterns inside each embedding region, Local Ternary Pattern (LTP) are then effectively exploited to locate the stable embedding positions inside embedding patterns in which the secret bits are embedded in. Finally, to make the scheme being robust against document rotation caused by distortion of printing and scanning process, Hough transform is applied to compute the rotation angle for restoring rotated document to original direction. Besides, repetition code and other improved methods are implemented to possibly enhance the accuracy of extracted secret data. The proposed steganography scheme in spatial domain is capable of detecting embedded data without any references and resisting to common image processing distortion.","PeriodicalId":433676,"journal":{"name":"2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116896170","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 7

LSDE: Levenshtein Space Deep Embedding for Query-by-String Word Spotting 基于字符串查询的Levenshtein空间深度嵌入

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) Pub Date : 2017-11-01 DOI: 10.1109/ICDAR.2017.88

L. G. I. Bigorda, Marçal Rusiñol, Dimosthenis Karatzas

引用次数: 25

Are Multidimensional Recurrent Layers Really Necessary for Handwritten Text Recognition? 手写文本识别真的需要多维循环层吗?

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) Pub Date : 2017-11-01 DOI: 10.1109/ICDAR.2017.20

J. Puigcerver

引用次数: 205

Compact and Efficient WFST-Based Decoders for Handwriting Recognition 紧凑高效的基于wfst的手写体识别解码器

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) Pub Date : 2017-11-01 DOI: 10.1109/ICDAR.2017.32

Meng Cai, Qiang Huo

引用次数: 9

Bag of Local Convolutional Triplets for Script Identification in Scene Text 基于局部卷积三联体的场景文本脚本识别

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) Pub Date : 2017-11-01 DOI: 10.1109/ICDAR.2017.68

Jan Zdenek, Hideki Nakayama

引用次数: 10

Text Proposals Based on Windowed Maximally Stable Extremal Region for Scene Text Detection 基于窗口最大稳定极值区域的场景文本检测文本建议

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) Pub Date : 2017-11-01 DOI: 10.1109/ICDAR.2017.69

Feng Su, Wenjun Ding, Lan Wang, Susu Shan, Hailiang Xu

引用次数: 2

ICDAR2017 Robust Reading Challenge on Omnidirectional Video ICDAR2017全向视频鲁棒阅读挑战

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) Pub Date : 2017-11-01 DOI: 10.1109/ICDAR.2017.236

M. Iwamura, Naoyuki Morimoto, Keishi Tainaka, Dena Bazazian, L. G. I. Bigorda, Dimosthenis Karatzas

{"title":"ICDAR2017 Robust Reading Challenge on Omnidirectional Video","authors":"M. Iwamura, Naoyuki Morimoto, Keishi Tainaka, Dena Bazazian, L. G. I. Bigorda, Dimosthenis Karatzas","doi":"10.1109/ICDAR.2017.236","DOIUrl":"https://doi.org/10.1109/ICDAR.2017.236","url":null,"abstract":"Results of ICDAR 2017 Robust Reading Challenge on Omnidirectional Video are presented. This competition uses Downtown Osaka Scene Text (DOST) Dataset that was captured in Osaka, Japan with an omnidirectional camera. Hence, it consists of sequential images (videos) of different view angles. Regarding the sequential images as videos (video mode), two tasks of localisation and end-to-end recognition are prepared. Regarding them as a set of still images (still image mode), three tasks of localisation, cropped word recognition and end-to-end recognition are prepared. As the dataset has been captured in Japan, the dataset contains Japanese text but also include text consisting of alphanumeric characters (Latin text). Hence, a submitted result for each task is evaluated in three ways: using Japanese only ground truth (GT), using Latin only GT and using combined GTs of both. Finally, by the submission deadline, we have received two submissions in the text localisation task of the still image mode. We intend to continue the competition in the open mode. Expecting further submissions, in this report we provide baseline results in all the tasks in addition to the submissions from the community.","PeriodicalId":433676,"journal":{"name":"2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126135475","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 17

Extremely Sparse Deep Learning Using Inception Modules with Dropfilters 极其稀疏的深度学习使用初始模块与Dropfilters

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) Pub Date : 2017-11-01 DOI: 10.1109/ICDAR.2017.80

Woo-Young Kang, Kyung-Wha Park, Byoung-Tak Zhang

引用次数: 0