Filters








244,134 Hits in 5.7 sec

End-to-End Text Classification via Image-based Embedding using Character-level Networks [article]

Shunsuke Kitada, Ryunosuke Kotani, Hitoshi Iyatomi
2018 arXiv   pre-print
The proposed CE-CLCNN is an end-to-end learning model and has an image-based character encoder, i.e. the CE-CLCNN handles each character in the target document as an image.  ...  open document classification tasks.  ...  [15] explicitly learned to preserve character shape features by CAE, but our CE-CLCNN does not explicitly learn character representation that preserves the shape Fig. 3 .  ... 
arXiv:1810.03595v2 fatcat:43yhkxnnanh5nfh2xrzbuzje44

SF-CNN: Deep Text Classification and Retrieval for Text Documents

R. Sarasu, K. K. Thyagharajan, N. R. Shanker
2023 Intelligent Automation and Soft Computing  
for retrieving correct text documents.  ...  Traditional deep learning methods such as Convolutional Neural Network and Recurrent Neural Network never use semantic representation for bag-of-words.  ...  The proposed SF-CNN method enhances the semantic features for classifying and retrieving research documents better than traditional methods.  ... 
doi:10.32604/iasc.2023.027429 fatcat:r2czwj5p6jdntkr3lgkp23erma

Discretization based learning approach to information retrieval

Dmitri Roussinov, Weiguo Fan, Fernando A. Das Neves
2005 Proceedings of the 14th ACM international conference on Information and knowledge management - CIKM '05  
We approached the problem as learning how to order documents by estimated relevance with respect to a user query.  ...  For this, we have designed a representation scheme, which is based on the discrete representation of the local (lw) and global (gw) weighting functions, thus is capable of reproducing and enhancing the  ...  Figure 1 . 1 Learning local weighting for various Figure 3 . 3 Learned optimal shape of local weighting. Figure 4 . 4 Learned optimal shape of global weighting G(t).  ... 
doi:10.1145/1099554.1099647 dblp:conf/cikm/RoussinovFN05a fatcat:ziy7zvhobjesxl2ze3uz5txgku

F-ratio Based Weighted Feature Extraction for Similar Shape Character Recognition

Tetsushi Wakabayashi, Umapada Pal, Fumitaka Kimura, Yasuji Miyake
2009 2009 10th International Conference on Document Analysis and Recognition  
This weighting scheme enhances the feature elements that belongs to the distinguishable portions of the similar shaped characters and reduces the feature elements of the common portion of the characters  ...  Fratio modifies the feature vector of two similar shape characters by weighting the feature elements.  ...  F-ratio is calculated from feature vectors belong to the similar shaped character classes and enhanced the feature vector for better recognition.  ... 
doi:10.1109/icdar.2009.197 dblp:conf/icdar/WakabayashiPKM09 fatcat:tsbvmggm7jcn5ghjnqbtnecu7q

Supplementary document for Unsupervised Hyperspectral Stimulated Raman Microscopy Image Enhancement: Denoising and Segmentation via One-Shot Deep Learning - 5472694.pdf

pedram Abdolghader, Andrew Ridsdale, Tassos Grammatikopoulos, Gavin Resch, François Légaré, Albert Stolow, Adrian Pegoraro, Isaac Tamblyn
2021 figshare.com  
Unsupervised Hyperspectral Image Enhancement, Segmentation and De-Noising in Stimulated Raman Microscopy: supplemental document 1.  ...  S4 The spatial distribution of the PSNR for FOV1 for the data presented inFig. 5for (a) Input-GT, (b) SHRED-GT, and (c) UHRED-GT.  ... 
doi:10.6084/m9.figshare.16705603.v1 fatcat:safluluck5a2fh4mh3esmt4o44

CUTIE: Learning to Understand Documents with Convolutional Universal Text Information Extractor [article]

Xiaohui Zhao, Endi Niu, Zhuo Wu, Xiaoguang Wang
2019 arXiv   pre-print
To avoid designing expert rules for each specific type of document, some published works attempt to tackle the problem by learning a model to explore the semantic context in text sequences based on the  ...  Extracting key information from documents, such as receipts or invoices, and preserving the interested texts to structured data is crucial in the document-intensive streamline processes of office automation  ...  Furthermore, to enhance the capability of CUTIE to better handle documents with different layouts, we augment the grid data to shapes with different rows and columns by random sampling a Gaussian distribution  ... 
arXiv:1903.12363v4 fatcat:ra73l3owrzftngnuohw5ftnkhy

The Role and Utilization of CNN in Automatic Logo Based Document Image Retrieval Methods

Raveendra K, R Vinoth Kanna
2018 International Journal of Engineering & Technology  
Automatic logo based document image retrieval process is an essential and mostly used method in the feature extraction applications.  ...  The main objective of this paper is to effectively utilize the CNN in the process of automatic logo based document image retrieval methods.  ...  R.Vinoth Kanna sir for his continuous supportive encouragement in my research work done so far including this paper.  ... 
doi:10.14419/ijet.v7i3.1.16786 fatcat:i7rwnytolffarejeiyfzankf3u

Enhanced visual statistical learning in adults with autism

Matthew E. Roser, Richard N. Aslin, Rebecca McKenzie, Daniel Zahra, József Fiser
2015 Neuropsychology  
Conclusions: These results extend previous observations of visuospatial enhancement in ASD into the domain of learning, and suggest that enhanced visual statistical learning may have arisen from a sustained  ...  bias to attend to local details in complex arrays of visual features.  ...  For example, a circular-shaped clock and a wheel share a feature ("roundness"), but it is a defining feature only for the wheel.  ... 
doi:10.1037/neu0000137 pmid:25151115 pmcid:PMC4340818 fatcat:cxblkrea35btbdk5if73bkdyle

How Does Learning Impact Development in Infancy? The Case of Perceptual Organization

Ramesh S. Bhatt, Paul C. Quinn
2010 Infancy  
The proposed framework is an attempt to account for this process in the domain of perception.  ...  However, other processes are not readily evident in young infants, and their development involves perceptual learning.  ...  For instance, being exposed to correlations between two features (say, shape and color: shape A being red always and shape B being blue always) in the context of other varying features (say, size: shapes  ... 
doi:10.1111/j.1532-7078.2010.00048.x pmid:21572570 pmcid:PMC3092381 fatcat:jfbssft5rvctrhznsrka5su5xa

HANDWRITTEN DEVANAGARI VOWEL RECOGNITION USING ARTIFICIAL NEURAL NETWORK

P E Ajmire
2017 International Journal of Advanced Research in Computer Science  
Human being is doing this task while learning characters in the childhood. But the same task for machine is much complex.  ...  This research work proposes new approaches for extracting features in context of Handwritten Devanagari Vowels recognition. For classification technique Artificial Network is used.  ...  HOG features describe the shape of the image by the distribution of intensity gradients or edge directions. A HOG feature vector represents local shape of an object [12, 13] .  ... 
doi:10.26483/ijarcs.v8i7.4560 fatcat:2bbqhg72erd3tfhid4sgsurzka

Document Image Retrieval Based on Keyword Spotting Using Relevance Feedback

M. Keyvanpour
2013 International Journal of Engineering  
Keyword Spotting is a well-known method in document image retrieval which is based on query word image.  ...  In this paper, a document image retrieval system based on keyword spotting and relevance feedback is presented.  ...  [15] proposed a new feedback approach with progressive learning capability combined with a novel method for feature subspace extraction.  ... 
doi:10.5829/idosi.ije.2014.27.01a.02 fatcat:nqqnogjecngexol6wdwb3h3rg4

A brief review of document image retrieval methods: Recent advances

Fahimeh Alaei, Alireza Alaei, Michael Blumenstein, Umapada Pal
2016 2016 International Joint Conference on Neural Networks (IJCNN)  
This paper provides an overview of the methods which have been applied for document image retrieval over recent years.  ...  Many techniques have been developed to provide an efficient and effective way for retrieving and organizing these document images in the literature.  ...  The indexing/learning methods are applied to train a classifier or knowledge-based method for some given documents.  ... 
doi:10.1109/ijcnn.2016.7727648 dblp:conf/ijcnn/AlaeiABP16 fatcat:5tzfmk55r5hmpa3tnhcj3chuji

An enhanced binarization framework for degraded historical document images

Wei Xiong, Lei Zhou, Ling Yue, Lirong Li, Song Wang
2021 EURASIP Journal on Image and Video Processing  
It uses a disk-shaped structuring element, whose radius is computed by the minimum entropy-based stroke width transform (SWT).  ...  AbstractBinarization plays an important role in document analysis and recognition (DAR) systems.  ...  First, we can improve the contrast between text and background by using machine learning or deep learning techniques to effectively achieve degraded document image enhancement in the preprocessing stage  ... 
doi:10.1186/s13640-021-00556-4 fatcat:aawmkhjf3ngs3bb3zbyyi4xz5q

A Framework for Content Sequencing from Junior to Senior Mathematics Curriculum

Musarurwa David Chinofunga, Philemon Chigeza, Subhashni Taylor
2022 Eurasia Journal of Mathematics, Science and Technology Education  
a tool for sequencing the mathematics content.  ...  Planning templates and samples are available to schools; however, it is imperative for teachers to understand the processes that underpin planning.  ...  Shapes and intercepts, asymptotes shapes and behavior and features, center and radii can all be brought under features of graphs. 2.  ... 
doi:10.29333/ejmste/11930 fatcat:fymdxhhnyzew7otuwl7ydudsu4

Image analysis for digital media applications

Hong Yan
2001 IEEE Computer Graphics and Applications  
Acknowledgment Our work on cartoon image analysis, handwriting recognition, and document image compression is supported by several grants from the Australian Research Council.  ...  detection 1 Shape from shading: finding 3D shapes from 2D images Relaxation labeling: object matching 13 1 Example of image enhancement: (a) the original color image and (b) the enhanced image.  ...  For example, we can achieve a high recognition rate for well-isolated characters by integrating several classifiers. 9 In research, useful ideas can be learned from a different field.  ... 
doi:10.1109/38.895126 fatcat:pfhnkx3zyrhtpmpwchb7dhg4ym
« Previous Showing results 1 — 15 out of 244,134 results