Invited paper Printed Thai character recognition using topological properties method

In this paper, we address the problem of brain tumor classification from radiology and histopathology data. A coarse-to-fine classification approach is adopted using a combination of deep features and Graph Convolution Network (GCN). As a first coarse step, we use 3D CNN to detect Glioblastoma from MRI images. In order to infer about Astrocytoma and Oligodendroglioma, Whole Slide Images (WSI) are employed in the second stage. During this fine classification stage, 2D CNN features are extracted at two different (global and local) magnification levels. A graph is constructed with nodes in the space of concatenated global and local features. Edges are constructed from feature similarity and graph topology. Finally, GCN is used with normalized graph Laplacian to ensure better relation-aware-representation leading to more accurate classification. Experimental comparisons on the CPM-RadPath2020 challenge dataset clearly demonstrate the state-of-the-art performance of our proposed strategy. The code implementation is available at https://github.com/arijitde92/BrainTumorClassification.

A number of strategies have been used to include spatial and topological properties in the image segmentation stage. It is generally accepted that grouping of nearby pixels by modelling neighbourhood relationships as (a, b) connected graphs may lead to meaningful image objects. In such approach, however, topological concepts may suffer from ambiguity since image elements (pixels) are two dimensional entities. This paper evaluates whether an alternative representation of digital images based both on Cartesian complexes and oriented matroids may improve multispectral image segmentation by enforcing topological and geometric properties and then be used in the classification stage. A conceptual model is defined, using Cartesian complexes, in order to link combinatorial properties of axiomatic locally finite spaces and their associated oriented matroids for involving topological properties. The proposed approach uses a layered architecture going from a physical level, going next through logical geospatial abstraction level and then through the Cartesian complex logical level. Additionally, there is a layer of oriented matroids composed by conceptual elements in terms of combinatorics for encoding relevant features to multispectral image segmentation. First, it is conducted an edge detection task, next an probability contour map using a Cartesian complex space rather than the conventional image space and finally, an image classification using random forest method. A computational solution including several components was developed using a framework for parallel computing. The performance of this solution was assessed using a small subset of GEOBIA2016 benchmark dataset. It is shown that the usage of a partial implementation of Cartesian complexes and associated oriented matroids is computationally but does not increase classification accuracy.

Traditionally, Thai characters are composed of circle, zigzag line, curve, and head. However, many new Thai fonts, which are now gaining in popularity, do not follow the traditional writing rule; the head has been omitted from the characters. Without the head, it is very difficult to segregate the characters. Even the best commercial Thai OCR software has difficulty in recognizing this kind of character. Therefore, the hierarchical cross-correlation ARTMAP is proposed in this paper to recognize the no-head Thai characters

Simultaneous Localization and Mapping (SLAM) is a fundamental problem in the field of robotics which concerns mapping an environment or space while simultaneously localizing within this map. Given that one of the major goals of robotics is to perform tasks commonly performed by humans, we argue that SLAM methods should be cognitively adequate; that is, they should model the same properties of a space as the human cognition models. Topological properties are considered the most fundamental of those modelled by the human cognition. Therefore in order to achieve cognitive adequacy such properties must be modelled explicitly. Research in the domain of spatial cognition has demonstrated that the topological properties modelled by the human cognition can be quantified using the Egenhofer Nine-Intersection Model (9-IM). In this work we propose a conceptual SLAM method which models the same properties as the 9-IM. Relative to existing topological SLAM methods, which model a single topological property of connectivity between locations, this method achieves a stronger degree of cognitive adequacy.

The idea that one could prove by analytic methods the topological invariance of characteristic classes was suggested in an influential lecture by Singer. The concrete realization of this, in the case of rational Pontrjagin classes, was the result of works by Sullivan, who showed that topological manifolds have essentially unique Lipschitz structures, and Teleman, who showed that Lipschitz manifolds are the type of object for which one can analyze the signature operator. (A refinement of the same method proves the KO[l/2] orientability of topological bundles.) Much of this can be done for manifolds with group actions. The main implication is that for certain sorts of topological group actions (including all smooth and PL actions) one can construct an signature operator whose class in equivariant if-homology is a topological invariant. This class satisfies the G-signature theorem. The topological method has implications for various precise classification problems (e.g. semifree topological actions on the sphere), but here we shall present only the consequences of these Lipschitzian ideas: (A) Topologically locally linear G-manifolds for odd-order G have KOG[1/2] orientations (Madsen-Rothenberg). For all finite G there are Lipschitz structures, but the signature class is a zero divisor for even-order groups. The same method produces, upon application of the localization theorem, Atiyah-Singer classes for some actions with nonmanifold fixed set (Weinberger, extending earlier joint work with Cappell). (B) For odd-order groups, topological and linear conjugacy of linear representations are equivalent (Hsiang-Pardon, Madsen-Rothenberg). This is a consequence of (A). Alternatively one can return to the original argument of Atiyah-Bott-Milnor for semifree representations. A stronger result is true: the Atiyah-Bott local numbers associated to representations agree for elements that have discrete fixed sets in topologically equivalent representations. This distinguishes many pairs of representations of even-order groups. (C) R Top(G) = R Lip(G) for all finite G, where R Cat is the Grothendieck group of representations up to Cat conjugacy (Weinberger, after inverting 2). This implies that there is no analytic definition of Reidemeister torsion that works well in the Lipschitz setting. The reason is that, following

In this paper; we introduce a system of automatic recognition of characters based on the Random Forest Method in non-constrictive pictures that are stemmed from the terminals Mobile phone. After doing some pretreatments on the picture, the text is segmented into lines and then into characters. In the stage of characteristics extraction, we are representing the input data into the vector of primitives of the zoning types, of diagonal, horizontal and of the Zernike moment. These characteristics are linked to pixels' densities and they are extracted on binary pictures. In the classification stage, we examine four classification methods with two different classifiers types namely the multi-layer perceptron (MLP) and the Random Forest method. After some checking tests, the system of learning and recognition which is based on the Random Forest has shown a good performance on a basis of 100 models of pictures.

This paper presents a system for off-line handwritten numeral recognition based on topological properties of the digits. The first step in recognition algorithm is graph evaluation, obtained from the RLC of the digit, the second is measuring the geometrical properties of the elements of the graph and classification based on that measures. Algorithm was trained and tested on CEDAR database and it achieved correct recognition rate of 99,74%.

Over the years, face recognition has been the research topic that has attracted many researchers around the world. One of the most significant applications of face recognition is the access control system. The access control system allows authorized persons to enter or exit certain or restricted areas. As a result, it will increase the security situation without over-investment in staff security. The access information can be the identification, time, and location, etc. It can be used to carry out human resource management tasks such as attendance and inspection of employees in a more fair and transparent manner. Although face recognition has been widely used in access control systems because of its better accuracy and convenience without requiring too much user cooperation, the 2D-based face recognition systems also retain many limitations due to the variations in pose and illumination. By analyzing facial geometries, 3D facial recognition systems can theoretically overcome the disadvantages of prior 2D methods and improve robustness in different working conditions. In this paper, we propose the 3D facial recognition algorithm for use in an access control system. The proposed algorithm includes the preprocessing, feature extraction, and classification stages. The application of the proposed access control system is the automatic sliding door, the controller of the system, the web-based monitoring, control, and storage of data.

In the paper, to handle the problem of traditional bag-of-features model ignoring the spatial relationship of local features in human action recognition, we proposed a Multiscale Spatial Position Coding under Locality Constraint method.Specifically, to describe this spatial relationship, we proposed a mixed feature combining motion feature and multi-spatial-scale configuration.To utilize temporal information between features, sub spatial-temporal-volumes are built.Next, the pooled features of sub-STVs are obtained via max-pooling method.In classification stage, the Locality-Constrained Group Sparse Representation is adopted to utilize the intrinsic group information of the sub-STV features.The experimental results on the KTH, Weizmann, and UCF sports datasets show that our action recognition system outperforms the classical local ST feature-based recognition systems published recently.

The Dynamical Regime of Active Regions via the Concept of Persistent Homology

Arabic language can be used by native and non- native speakers; due to Arabic is the language of the holy book of Muslims. In this paper, Arabic phoneme recognition system is proposed based on Malay speakers. This system consists of three main stages. The first stage is noise reduction and it aims to enhance the phoneme signals by excluding the unvoiced signals and keep only the voiced signal. Wiener filter is adapted to accomplish this task. The second stage is based on Mel-Frequency Cepstral Coefficients method to extract a vector of features to represent each phoneme signal. Eventually, pattern recognition neural network is designed as recognizer. The proposed system produces sufficient outcomes with 20 hidden neurons. Keywords—Arabic; Malay; pattern recognition; Wiener; Mel- Frequency Cepstral Coefficients I. INTRODUCTION It is well known the wide range of Artificial Neural Networks (ANN) application in speech recognition, financial, telecommunications, electronics and medical. As for speech recognition system for Arabic language, (1) & (2) have addressed the implementation of recognition system based on Arabic native speakers. In general, neural networks (NNs) for phonemes recognition are divided into several categories for instance Very Large Scale Integration (VLSI) NN that can be divided into digital and analogue with digital NNs being more compatible with feed-forward neural networks while analogue NNs are found to be more successful with recurrent NNs (3). From Automatic Speech System (ASR) point of view, NNs for recognition purpose can be divided into two main category namely conventional neural networks such as Multilayer Perceptron (MLP) and Radial Basis Function (RBF) and secondly Recurrent Neural Networks (RNN). The first category of NNs was implemented as pattern classifiers and proven capable for recognition purpose but not at par as compared to Hidden Markov Model (HMM) performance (2). At present, HMM has been proven to be the most successful approach in ASR research area until recently the possibility to hybrid both HMM and ANN (4) & (5). In general, speech recognition systems consist of three stages specifically noise reduction, feature extraction and classification. The aim of noise reduction is to reduce the level of noise in the speech signal and make it more compatible for the ASR. As for feature extraction, this process attempted to extract the most relevant features related to the phoneme signal. The extracted features acted as input features of the NN during training and testing. Finally, the purpose of classification process is to arrange the most similar and related features in categories based on the training phase. Basically, networks training database are being observed and updated during training session of the networks, involving the weights and biases arguments. Networks with lower errors can be considered as better networks (6). Hence, this paper deems to explore further the ability of NN for Arabic phonemes recognition spoken by Malay individuals. This paper is arranged as following; related work is discussed in section II followed by section III that discussed method to be implemented based on Wiener filter, Mel-Frequency Cepstral Coefficients and pattern recognition neural network. Section IV elaborated experimental analysis and results and followed by Conclusion section.

In this paper; we introduce a system of automatic recognition of Video Text Amazigh based on the Random Forest. After doing some pretreatments on the video and picture, the text is segmented into lines and then into characters. In the stage of characteristics extraction, we are representing the input data into the vector of primitives. These characteristics are linked to pixels’ densities and they are extracted on binary pictures. In the classification stage, we examine four classification methods with two different classifiers types namely the convolutional neural network (CNN) and the Random Forest method. We carried out the experiments with a database containing 3300 samples collected from different writers. The experimental results show that our proposed OCR system is very efficient and provides good recognition accuracy rate of handwriting characters images acquired via Video camera phone.

Voice recognition, command and control systems have played a vital role to facilitate our modern lifestyle and routine operations through hands-free applications. Working people are busy with their routines, simple tasks such as to on/off electrical appliances at home are often overlooked leads to wastage of energy. Besides, the elderly and handicapped often face difficulties to control the electrical appliances due to their physical disability. To overcome, this project aims to develop a voice command intelligent system (VCIS) as an aid to these routine tasks such as to switch on/off a lamp, control a motor and control other modes of operation of home appliances using voice interface. In particular, VCIS involves several development processes such as database collection, pre-processing, feature extraction algorithms and classification modeling for application in smart home. This paper proposes three hierarchical stages in VCIS. The experimental results show that Mel-frequency cepstral coefficients (MFCC) surpassed linear prediction coefficients (LPC) to correlate with the voice commands in all classification stages using artificial neural network. The classification rates were successfully accomplished, 99.12% (MFCC) and 95.23% (LPC). In conclusion, the proposed VCIS can be employed as an efficient method for better quality of life.

Breast cancer is the second most common type of cancer in the world. Several computer-aided detection and diagnosis systems have been used to assist health experts and to indicate suspect areas that would be difficult to perceive by the human eye; this approach has aided in the detection and diagnosis of cancer. The present work proposes a method for the automatic detection of masses in digital mammograms by using quality threshold (QT), a correlogram function, and the support vector machine (SVM). This methodology comprises the following steps: The first step is to perform preprocessing with a low-pass filter, which increases the scale of the contrast, and the next step is to use an enhancement to the wavelet transform with a linear function. After the preprocessing is segmentation using QT; then, we perform post-processing, which involves the selection of the best mass candidates. This step is performed by analyzing the shape descriptors through the SVM. For the stage that involves the extraction of texture features, we used Haralick descriptors and a correlogram function. In the classification stage, the SVM was again used for training, validation, and final test. The results were as follows: sensitivity 92.31 %, specificity 82.2 %, accuracy 83.53 %, mean rate of false positives per image 1.12, and area under the receiver operating characteristic (ROC) curve 0.8033. Breast cancer is notable for presenting the highest mortality rate in addition to one of the smallest survival rates after diagnosis. An early diagnosis means a considerable increase in the survival chance of the patients. The methodology proposed herein contributes to the early diagnosis and survival rate and, thus, proves to be a useful tool for specialists who attempt to anticipate the detection of masses.

Interval routing was introduced to reduce the size of routing tables: a router finds the direction where to forward a message by determining which interval contains the destination address of the message, each interval being associated to one particular direction. This way of implementing a routing function is quite attractive but very little is known about the topological properties that must satisfy a network to support an interval routing function with particular constraints (shortest paths, limited number of intervals associated to each direction, etc.). In this paper we investigate the study of the interval routing functions. In particular, we characterize the set of networks which support a linear or a linear strict interval routing function with only one interval per direction. We also derive practical tools to measure the efficiency of an interval routing function (number of intervals, length of the paths, etc.), and we describe large classes of networks which support optimal (linear) interval routing functions. Finally, we derive the main properties satisfied by the popular networks used to interconnect processors in a distributed memory parallel computer.