scholarly journals Optimisation of the Largest Annotated Tibetan Corpus Combining Rule-based, Memory-based, and Deep-learning Methods

2021 ◽  
Vol 20 (1) ◽  
pp. 1-11
Author(s):  
Marieke Meelen ◽  
Élie Roux ◽  
Nathan Hill
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Rule Based ◽  
Plain Text ◽  
Resource Center ◽  
Digital Resource ◽  
Network Methods ◽  
Supervised Methods ◽  
Tibetan Studies ◽  
Automatic Pipeline

This article presents a pipeline that converts collections of Tibetan documents in plain text or XML into a fully segmented and POS-tagged corpus. We apply the pipeline to the large extent collection of the Buddhist Digital Resource Center. The semi-supervised methods presented here not only result in a new and improved version of the largest annotated Tibetan corpus to date, the integration of rule-based, memory-based, and neural-network methods also serves as a good example of how to overcome challenges of under-researched languages. The end-to-end accuracy of our entire automatic pipeline of 91.99% is high enough to make the resulting corpus a useful resource for both linguists and scholars of Tibetan studies.

scholarly journals Defect Detection in CK45 Steel Structures through C-scan Images Using Deep Learning Method

2021 ◽  
Vol 3 (1) ◽  
Author(s):  
Navid Moshtaghi Yazdani
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Measurement System ◽  
Processing System ◽  
Steel Structures ◽  
Learning Method ◽  
Metallic Structures ◽  
Defect Profile ◽  
Network Methods ◽  
Deep Convolution Neural Network

In the present paper, a method for reliable estimation of defect profile in CK45 steel structures is presented using an eddy current testing based measurement system and post-processing system based on deep learning technique. So a deep learning method is used to determine the defect characteristics in metallic structures by magnetic field C-scan images obtained by an anisotropic magneto-resistive sensor. Having designed and adjusting the deep convolution neural network and applied it to C-scan images obtained from the measurement system, the performance of deep learning method proposed is compared with conventional artificial neural network methods such as multilayer perceptron and radial basis function on a number of metallic specimens with different defects. The results confirm the superiority of the proposed method for characterizing defects compared to other classical training-oriented methods.

Comparison of Fundamental Learning Strategies for CBIR using Deep Learning Methods

2021 ◽  
Vol 14 ◽  
Author(s):  
Meenakshi Garg ◽  
Manisha Malhotra ◽  
Harpal Singh
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Learning Strategies ◽  
Deep Neural Network ◽  
Fuzzy Classification ◽  
Digital Archive ◽  
Large Database ◽  
Fuzzy Classifier ◽  
Neuro Fuzzy ◽  
Network Methods

Background: Photo retrieval based on contents is primarily used to retrieve photographs from a broad database. CBIR, also named "search by image," is an al-lowing technology that handles computerized images by its recognizable attributes. Methods: In other words, CBIR is a method for recovery of images that does not rely on annotations or keywords but on the characteristics of the images directly taken from the pictures. CBIR systems rely on the use of machine display methods in broad datasets for the image retrieval issue. The CBIR technology is the retrieval from a cluster of photos or archive of the most visually similar photographs to a particular query file.It is really useful for scanning photos, medical research etc. in other fields such as photography. It may be hard to visually find the images by inserting the metadata or keywords into a large database and cannot catch the keyword for identifying this image. CBIR allows the extraction of similar photographs from a digital archive with no labeling of photographs. The Deep Neural Network and Neuro-Fuzzy classification are contrasted in this article. They both have numerous findings and numerous tests to forecast the picture. Results: The analysis of the neuro-fuzzy and deep neural network methods we suggest reveals that the precision is increased. Conclusion: Accuracy values for DNN and Neuro-Fuzzy Classifier process are74.6% and 75.4%. For the validity of the proposed process, the visual and qualitative findings are provided.

scholarly journals NEW APPROACH FOR ONLINE ARABIC MANUSCRIPT RECOGNITION BY DEEP BELIEF NETWORK

2018 ◽  
Vol 58 (5) ◽  
pp. 297
Author(s):  
Benbakreti Samir ◽  
Aoued Boukelif
Keyword(s):  
Neural Network ◽  
Feature Extraction ◽  
Deep Learning ◽  
Success Rate ◽  
Character Recognition ◽  
Recognition Rate ◽  
Deep Belief Network ◽  
New Approach ◽  
Belief Network ◽  
Network Methods

In this paper, we present a neural approach for an unconstrained Arabic manuscript recognition using the online writing signal rather than images. First, we build the database which contains 2800 characters and 4800 words collected from 20 different handwritings. Thereafter, we will perform the pretreatment, feature extraction and classification phases, respectively. The use of a classical neural network methods has been beneficial for the character recognition, but revealed some limitations for the recognition rate of Arabic words. To remedy this, we used a deep learning through the Deep Belief Network (DBN) that resulted in a 97.08% success rate of recognition for Arabic words.

RANDOM NEURAL NETWORK METHODS AND DEEP LEARNING

2019 ◽  
pp. 1-31 ◽  
Author(s):  
Yonghua Yin
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Mathematical Model ◽  
Deep Learning ◽  
Energy Expenditure ◽  
Great Success ◽  
Integrate And Fire ◽  
Random Neural Network ◽  
Network Methods ◽  
Spiking Network

The random neural network (RNN) is a mathematical model for an “integrate and fire” spiking network that closely resembles the stochastic behavior of neurons in mammalian brains. Since its proposal in 1989, there have been numerous investigations into the RNN's applications and learning algorithms. Deep learning (DL) has achieved great success in machine learning. Recently, the properties of the RNN for DL have been investigated, in order to combine their power. Recent results demonstrate that the gap between RNNs and DL can be bridged and the DL tools based on the RNN are faster and can potentially be used with less energy expenditure than existing methods.

scholarly journals Classification of Indramayu’s Mangoes Based on Shape Features Using Neural Network and Rule Based Programming

2017 ◽  
Vol 1 ◽  
pp. 245-248
Author(s):  
Sumarudin Sumarudin ◽  
Iryanto Iryanto ◽  
Eka Ismantohadi
Keyword(s):  
Neural Network ◽  
Image Processing ◽  
Deep Learning ◽  
Learning Approach ◽  
Shape Features ◽  
Rule Based ◽  
General Classification ◽  
Result Show ◽  
Better Than

Object classification using image processing simplifies the process. Many approaches have been used to classify the object. In general, classification of mangoes uses image of leaves. In this research, we do a slightly different approach using image of mango itself. Here, two kinds of method are used to classify the object.  Implementations of deep learning using neural network and rule based programming are used in the process. Comparative study of the methods are presented in the article. Our result show that accuracy of deep learning approach is better than the rule based programming. The accuracy is 80% and 8% for neural network and rule based programming, respectively.

scholarly journals The Progress of Medical Image Semantic Segmentation Methods for Application in COVID-19 Detection

2021 ◽  
Vol 2021 ◽  
pp. 1-26
Author(s):  
Amin Valizadeh ◽  
Morteza Shariatee
Keyword(s):  
Neural Network ◽  
Traditional Method ◽  
Medical Image ◽  
Deep Neural Network ◽  
Sampling Methods ◽  
Semantic Segmentation ◽  
Intelligent Transportation ◽  
Segmentation Methods ◽  
Network Methods ◽  
Supervised Methods

Image medical semantic segmentation has been employed in various areas, including medical imaging, computer vision, and intelligent transportation. In this study, the method of semantic segmenting images is split into two sections: the method of the deep neural network and previous traditional method. The traditional method and the published dataset for segmentation are reviewed in the first step. The presented aspects, including all-convolution network, sampling methods, FCN connector with CRF methods, extended convolutional neural network methods, improvements in network structure, pyramid methods, multistage and multifeature methods, supervised methods, semiregulatory methods, and nonregulatory methods, are then thoroughly explored in current methods based on the deep neural network. Finally, a general conclusion on the use of developed advances based on deep neural network concepts in semantic segmentation is presented.

scholarly journals Deep Semantic Compliance Advisor for Unstructured Document Compliance Checking

2020 ◽  
Author(s):  
Honglei Guo ◽  
Bang An ◽  
Zhili Guo ◽  
Zhong Su
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Natural Language ◽  
Semantic Relatedness ◽  
Rule Based ◽  
Legal Professionals ◽  
Detection Model ◽  
Semantic Complexity ◽  
Compliance Checking ◽  
Multi Level

Unstructured document compliance checking is always a big challenge for banks since huge amounts of contracts and regulations written in natural language require professionals' interpretation and judgment. Traditional rule-based or keyword-based methods cannot precisely characterize the deep semantic distribution in the unstructured document semantic compliance checking due to the semantic complexity of contracts and regulations. Deep Semantic Compliance Advisor (DSCA) is an unstructured document compliance checking platform which provides multi-level semantic comparison by deep learning algorithms. In the statement-level semantic comparison, a Graph Neural Network (GNN) based syntactic sentence encoder is proposed to capture the complicate syntactic and semantic clues of the statement sentences. This GNN-based encoder outperforms existing syntactic sentence encoders in deep semantic comparison and is more beneficial for long sentences. In the clause-level semantic comparison, an attention-based semantic relatedness detection model is applied to find the most relevant legal clauses. DSCA significantly enhances the productivity of legal professionals in the unstructured document compliance checking for banks.

scholarly journals Fuzzy Superpixels Based Semi-Supervised Similarity-Constrained CNN for PolSAR Image Classification

2020 ◽  
Vol 12 (10) ◽  
pp. 1694 ◽  
Author(s):  
Yuwei Guo ◽  
Zhuangzhuang Sun ◽  
Rong Qu ◽  
Licheng Jiao ◽  
Fang Liu ◽  
...  
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Computational Complexity ◽  
Convolutional Neural Network ◽  
Image Classification ◽  
Classification Performance ◽  
Unlabeled Data ◽  
Training Set ◽  
Supervised Methods ◽  
High Computational Complexity

Recently, deep learning has been highly successful in image classification. Labeling the PolSAR data, however, is time-consuming and laborious and in response semi-supervised deep learning has been increasingly investigated in PolSAR image classification. Semi-supervised deep learning methods for PolSAR image classification can be broadly divided into two categories, namely pixels-based methods and superpixels-based methods. Pixels-based semi-supervised methods are liable to be affected by speckle noises and have a relatively high computational complexity. Superpixels-based methods focus on the superpixels and ignore tiny detail-preserving represented by pixels. In this paper, a Fuzzy superpixels based Semi-supervised Similarity-constrained CNN (FS-SCNN) is proposed. To reduce the effect of speckle noises and preserve the details, FS-SCNN uses a fuzzy superpixels algorithm to segment an image into two parts, superpixels and undetermined pixels. Moreover, the fuzzy superpixels algorithm can also reduce the number of mixed superpixels and improve classification performance. To exploit unlabeled data effectively, we also propose a Similarity-constrained Convolutional Neural Network (SCNN) model to assign pseudo labels to unlabeled data. The final training set consists of the initial labeled data and these pseudo labeled data. Three PolSAR images are used to demonstrate the excellent classification performance of the FS-SCNN method with data of limited labels.

scholarly journals DETECTION OF DISEASE ON CORN PLANTS USING CONVOLUTIONAL NEURAL NETWORK METHODS

2019 ◽  
Vol 12 (1) ◽  
pp. 51 ◽  
Author(s):  
Ardi Hidayat ◽  
Ucuk Darusalam ◽  
Irmawati Irmawati
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Leaf Spot ◽  
Plant Disease ◽  
Gray Leaf Spot ◽  
Corn Plant ◽  
Interesting Issue ◽  
Network Methods ◽  
Corn Plants

Deep Learning is still an interesting issue and is still widely studied. In this study Deep Learning was used for the diagnosis of corn plant disease using the Convolutional Neural Network (CNN) method, with a total dataset of 3.854 images of diseases in corn plants, which consisted of three types of corn diseases namely Common Rust, Gray Leaf Spot, and Northern Leaf Blight. With an accuracy of 99%, in detecting disease in corn plants.

Sign in / Sign up

Export Citation Format

Share Document