Abstract: With the recent advance of the deep neural network, we observe new applications of NLP (natural language processing) and CV (computer vision) powered by deep neural networks for processing business documents. However, creating a real-world document processing system needs to integrate several NLP and CV tasks, rather than treating them separately. There is a need to have a unified approach for processing documents containing textual and graphical elements with rich formats, diverse layout arrangement, and distinct semantics. In this paper, a framework that fulfills this unified approach is presented. The framework includes a representation model definition for holding the information generated by various tasks and specifications defining the coordination between these tasks. The framework is a blueprint for building a system that can process documents with rich formats, styles, and multiple types of elements. The flexible and lightweight design of the framework can help build a system for diverse business scenarios, such as contract monitoring and reviewing.
Abstract: In pattern recognition applications the low level segmentation and the high level object recognition are generally considered as two separate steps. The paper presents a method that bridges the gap between the low and the high level object recognition. It is based on a Bayesian network representation and network propagation algorithm. At the low level it uses hierarchical structure of quadratic spline wavelet image bases. The method is demonstrated for a simple circuit diagram component identification problem.
Abstract: Most of the Question Answering systems
composed of three main modules: question processing,
document processing and answer processing. Question
processing module plays an important role in QA systems. If
this module doesn't work properly, it will make problems for
other sections. Moreover answer processing module is an
emerging topic in Question Answering, where these systems
are often required to rank and validate candidate answers.
These techniques aiming at finding short and precise answers
are often based on the semantic classification.
This paper discussed about a new model for question
answering which improved two main modules, question
processing and answer processing.
There are two important components which are the bases
of the question processing. First component is question
classification that specifies types of question and answer.
Second one is reformulation which converts the user's
question into an understandable question by QA system in a
specific domain. Answer processing module, consists of
candidate answer filtering, candidate answer ordering
components and also it has a validation section for interacting
with user. This module makes it more suitable to find exact
answer. In this paper we have described question and answer
processing modules with modeling, implementing and
evaluating the system. System implemented in two versions.
Results show that 'Version No.1' gave correct answer to 70%
of questions (30 correct answers to 50 asked questions) and
'version No.2' gave correct answers to 94% of questions (47
correct answers to 50 asked questions).
Abstract: Question answering (QA) aims at retrieving precise information from a large collection of documents. Most of the Question Answering systems composed of three main modules: question processing, document processing and answer processing. Question processing module plays an important role in QA systems to reformulate questions. Moreover answer processing module is an emerging topic in QA systems, where these systems are often required to rank and validate candidate answers. These techniques aiming at finding short and precise answers are often based on the semantic relations and co-occurrence keywords. This paper discussed about a new model for question answering which improved two main modules, question processing and answer processing which both affect on the evaluation of the system operations. There are two important components which are the bases of the question processing. First component is question classification that specifies types of question and answer. Second one is reformulation which converts the user's question into an understandable question by QA system in a specific domain. The objective of an Answer Validation task is thus to judge the correctness of an answer returned by a QA system, according to the text snippet given to support it. For validating answers we apply candidate answer filtering, candidate answer ranking and also it has a final validation section by user voting. Also this paper described new architecture of question and answer processing modules with modeling, implementing and evaluating the system. The system differs from most question answering systems in its answer validation model. This module makes it more suitable to find exact answer. Results show that, from total 50 asked questions, evaluation of the model, show 92% improving the decision of the system.
Abstract: Document image processing has become an
increasingly important technology in the automation of office
documentation tasks. During document scanning, skew is inevitably
introduced into the incoming document image. Since the algorithm
for layout analysis and character recognition are generally very
sensitive to the page skew. Hence, skew detection and correction in
document images are the critical steps before layout analysis. In this
paper, a novel skew detection method is presented for binary
document images. The method considered the some selected
characters of the text which may be subjected to thinning and Hough
transform to estimate skew angle accurately. Several experiments
have been conducted on various types of documents such as
documents containing English Documents, Journals, Text-Book,
Different Languages and Document with different fonts, Documents
with different resolutions, to reveal the robustness of the proposed
method. The experimental results revealed that the proposed method
is accurate compared to the results of well-known existing methods.
Abstract: This paper includes two novel techniques for skew
estimation of binary document images. These algorithms are based on
connected component analysis and Hough transform. Both these
methods focus on reducing the amount of input data provided to
Hough transform. In the first method, referred as word centroid
approach, the centroids of selected words are used for skew detection.
In the second method, referred as dilate & thin approach, the selected
characters are blocked and dilated to get word blocks and later
thinning is applied. The final image fed to Hough transform has the
thinned coordinates of word blocks in the image. The methods have
been successful in reducing the computational complexity of Hough
transform based skew estimation algorithms. Promising experimental
results are also provided to prove the effectiveness of the proposed
methods.
Abstract: In pattern recognition applications the low level
segmentation and the high level object recognition are generally
considered as two separate steps. The paper presents a method that
bridges the gap between the low and the high level object
recognition. It is based on a Bayesian network representation and
network propagation algorithm. At the low level it uses hierarchical
structure of quadratic spline wavelet image bases. The method is
demonstrated for a simple circuit diagram component identification
problem.
Abstract: In this paper, an efficient structural approach for
recognizing on-line handwritten digits is proposed. After reading
the digit from the user, the slope is estimated and normalized for
adjacent nodes. Based on the changing of signs of the slope values,
the primitives are identified and extracted. The names of these
primitives are represented by strings, and then a finite state
machine, which contains the grammars of the digits, is traced to
identify the digit. Finally, if there is any ambiguity, it will be
resolved. Experiments showed that this technique is flexible and
can achieve high recognition accuracy for the shapes of the digits
represented in this work.