| 书目名称 | Intelligent Data Engineering and Analytics | | 副标题 | Proceedings of the 9 | | 编辑 | Suresh Chandra Satapathy,Peter Peer,Anumoy Ghosh | | 视频video | | | 概述 | Presents research works in intelligent data engineering and analytics.Provides results of FICTA 2021 held at NIT Mizoram, Aizwal, Mizoram, India.Serves as a reference for researchers and practitioners | | 丛书名称 | Smart Innovation, Systems and Technologies | | 图书封面 |  | | 描述 | This book presents the proceedings of the 9th International Conference on Frontiers of Intelligent Computing: Theory and Applications (FICTA 2021), held at NIT Mizoram, Aizwal, Mizoram, India, during June 25 – 26, 2021. FICTA conference aims to bring together researchers, scientists, engineers, and practitioners to exchange their new ideas and experiences in the domain of intelligent computing theories with prospective applications to various engineering disciplines. .This volume covers broad areas of Intelligent Data Engineering and Analytics. The conference papers included herein presents both theoretical as well as practical aspects of data intensive computing, data mining, big data, knowledge management, intelligent data acquisition and processing from sensors, data communication networks protocols and architectures, etc. The volume will also serve as a knowledge centre for students of post-graduate level in various engineering disciplines. . | | 出版日期 | Conference proceedings 2022 | | 关键词 | Computational Intelligence; Artificial Intelligence; Data Mining and Knowledge Discovery; Big Data; Data | | 版次 | 1 | | doi | https://doi.org/10.1007/978-981-16-6624-7 | | isbn_softcover | 978-981-16-6626-1 | | isbn_ebook | 978-981-16-6624-7Series ISSN 2190-3018 Series E-ISSN 2190-3026 | | issn_series | 2190-3018 | | copyright | The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapor |
| 1 |
Front Matter |
|
|
Abstract
|
| 2 |
,Automated Flower Species Identification by Using Deep Convolution Neural Network, |
Shweta Bondre,Uma Yadav |
|
Abstract
In machine learning, image classification plays a very important role in demonstrating any image. Recognition of flower species is based on the geometry, texture, and form of different flowers in the past year. Now, nowadays, flower identification is widely used to recognize medicinal plant species. There are about 400,000 flowering plant species, and modern search engines have the mechanism to search and identify the image containing a flower, but due to millions of flower species worldwide, robustness is lacking. The method of machine learning with CNN is then used to classify the flower species in this proposed research work. With data, we will train the machine learning model, and if any unknown pattern is discovered, then the predictive model will predict the flower species by what it has been gained by the trained data. The built-in camera of the mobile phone will acquire the images of the flower species, and the flower image extraction function is performed using a pretrained network extraction of complex features.
|
| 3 |
,Information Retrieval for Cloud Forensics, |
Prasad Purnaye,Vrushali Kulkarni |
|
Abstract
The use of cloud computing and cloud-based services has increased sharply in the past decade. Regardless of many advantages, the extensive use of the cloud has also created a large attack platform, frequently exploited by cybercriminals, requiring real-time, automated detection and competent forensics tools for investigations. Evidence identification so far has been limited to performance studies on datasets that were created a long time ago and are not specific to cloud environments. In this paper, we introduce a novel dataset for cloud-specific evidence detection. The dataset has two categories: the monitoring database and the evidence database. The monitoring database has 43 features and 9610 records, whereas the evidence database has 360 memory dump files of around 280 GB which contain memory dumps of benign virtual machines and a hostile virtual machine. The dataset will be an important resource for evidence identification research in the cloud environment.
|
| 4 |
,Machine Translation System Combination with Enhanced Alignments Using Word Embeddings, |
Ch Ram Anirudh,Kavi Narayana Murthy |
|
Abstract
Machine Translation (MT) is a challenging problem and various techniques proposed for MT have their own strengths and weaknesses. Combining various MT systems has shown promising results. . is one such approach. In this work, we propose using word embeddings for aligning words from different hypotheses during confusion network generation. Our experiments, on English-Hindi language pair, have shown statistically significant improvement in BLEU scores, when compared to the baseline system combination. Four data-driven MT systems are combined, namely, a phrase based MT, hierarchical-phrase based MT, bi-directional recurrent neural network MT and transformer based MT. All of these have been trained on IIT Bombay English-Hindi parallel corpus.
|
| 5 |
,Geometry-Based Machining Feature Retrieval with Inductive Transfer Learning, |
N. S. Kamal,H. B. Barathi Ganesh,V. V. Sajith Variyar,V. Sowmya,K. P. Soman |
|
Abstract
Manufacturing industries have widely adopted the reuse of machine parts as a method to reduce costs and as a sustainable manufacturing practice. Identification of reusable features from the design of the parts and finding their similar features from the database is an important part of this process. In this project, with the help of fully convolutional geometric features, we are able to extract and learn the high-level semantic features from CAD models with inductive transfer learning. The extracted features are then compared with that of other CAD models from the database using Frobenius norm and identical features are retrieved. Later we passed the extracted features to a deep convolutional neural network with a spatial pyramid pooling layer and the performance of the feature retrieval increased significantly. It was evident from the results that the model could effectively capture the geometrical elements from machining features.
|
| 6 |
,Grapheme to Phoneme Conversion for Malayalam Speech Using Encoder-Decoder Architecture, |
R. Priyamvada,D. Govind,Vijay Krishna Menon,B. Premjith,K. P. Soman |
|
Abstract
The two key components of Automatic Speech Recognition (ASR) and Text-to-Speech (TTS) systems are language modeling and acoustic modeling. The language model generates a lexicon, which is a pronunciation dictionary. A lexicon can be created using a variety of approaches. For low-resource languages, rule-based methods are typically employed to build the lexicon. However, because the corpus is often tiny, this methodology does not account for all possible pronunciation variances. As a result, low-resource languages like Malayalam require a method for developing a comprehensive lexicon as the corpus grows. In this work, we explored deep learning based encoder-decoder models for grapheme-to-phoneme (G2P) conversion in Malayalam. Long Short-Term Memory (LSTM) and Bidirectional Long Short-Term Memory (BiLSTM) encoder models with varying embedding dimensions were used to create the encoder model. The performance of the deep learning models used for G2P conversion was measured using the Word Error Rate (WER) and Phoneme Error Rate (PER). With 1024 embedding dimensions, the encoder using the BiLSTM model had the maximum accuracy of 98.04% and the lowest PER of 2.57% at the phoneme level, an
|
| 7 |
,Usage of Blockchain Technology in e-Voting System Using Private Blockchain, |
Suman Majumder,Sangram Ray |
|
Abstract
In India, the conventional voting system consists of paper polling, electronic ballot system, and associated mechanical devices. However, for the improvement of conventional voting system, its scalability and accessibility from anywhere, a digital voting system can be built using the electronic polling devices like e-voting, mobile voting, and IoT-related centralized voting system. But the conventional system faces some limitations—fake voters, costing, quicker outcomes, constant Internet connectivity, secure user interface, snooping, and DoS attack. To remove those barriers, blockchain is introduced in e-voting applications that provide stability and anonymity of voters due to the use of Merkle tree and hashed confidential data and any changes can be detected if the hash value is changed and the message is conveyed immediately. In this scheme, we have proposed an e-voting application using private blockchain, practical Byzantine fault-tolerant (PBFT) non-competitive consensus algorithm, and ECC-based session generated by server.
|
| 8 |
,Bengali Visual Genome: A Multimodal Dataset for Machine Translation and Image Captioning, |
Arghyadeep Sen,Shantipriya Parida,Ketan Kotwal,Subhadarshi Panda,Ondřej Bojar,Satya Ranjan Dash |
|
Abstract
Multimodal machine translation (MMT) refers to the extraction of information from more than one modality aiming at performance improvement by utilizing information collected from the modalities other than pure text. The availability of multimodal datasets, particularly for Indian regional languages, is still limited, and thus, there is a need to build such datasets for regional languages to promote the state of MMT research. In this work, we describe the process of creation of the Bengali Visual Genome (BVG) dataset. The BVG is the first multimodal dataset consisting of text and images suitable for English-to-Bengali multimodal machine translation tasks and multimodal research. We also demonstrate the sample use-cases of machine translation and region-specific image captioning using the new BVG dataset. These results can be considered as the baseline for subsequent research.
|
| 9 |
,Deep Learning-Based Mosquito Species Detection Using Wingbeat Frequencies, |
Ayush Jhaveri,K. S. Sangwan,Vinod Maan,Dhiraj |
|
Abstract
The outbreak of mosquito-borne diseases such as malaria, dengue, chikungunya, Zika, yellow fever, and lymphatic filariasis has become a major threat to human existence. Hence, the elimination of harmful mosquito species has become a worldwide necessity. The techniques to reduce and eliminate these mosquito species require the monitoring of their populations in regions across the globe. This monitoring can be performed by automatic detection from the sounds of their wingbeats, which can be recorded in mosquito suction traps. In this paper, using the sounds emitted from their wingbeats, we explore the detection of the six most harmful mosquito species. From 279,566 wingbeat recordings in the wingbeat kaggle dataset, we balance the data across the six mosquito species using data augmentation techniques. With the use of state-of-the-art machine learning models, we achieve detection accuracies of up to 97%. These models can then be integrated with mosquito suction traps to form an efficient mosquito species detection system.
|
| 10 |
,Developments in Capsule Network Architecture: A Review, |
Sudarshan Kapadnis,Namita Tiwari,Meenu Chawla |
|
Abstract
Problems like image recognition, object detection, image segmentation need efficacious solution for computer vision. Traditionally, these problems are being solved by Deep Learning. Convolutional Neural Network. and Recurrent Neural Network models are used for computer vision tasks. However, CNNs have some drawbacks. They cannot recognize objects if same object is viewed from different viewpoints and deformed objects. Besides, CNNs require an immense amount of training data. Capsule networks are viewed as a new solution for computer vision problems. They are capable of solving the above-mentioned problems better than CNNs. Capsule networks have shown better accuracy in many computer vision applications. In this paper, we review the methodologies and architectures of existing implementations of capsule networks.
|
| 11 |
,Computer-Aided Segmentation of Polyps Using Mask R-CNN and Approach to Reduce False Positives, |
Saurabh Jha,Balaji Jagtap,Srijan Mazumdar,Saugata Sinha |
|
Abstract
Computer-aided detection and segmentation of polyps present inside colon are quite challenging due to the large variations of polyps in features like shape, texture, size, and color, and the presence of various polyp-like structures during colonoscopy. In this paper, we apply a mask region-based convolutional neural network (Mask R-CNN) approach for the detection and segmentation of polyps in the images obtained from colonoscopy videos. We propose an efficient method to reduce the false positives in the computer-aided detection system. To achieve this, we rigorously train our model by selecting non-polyp regions in the image which have high probability of getting detected as a polyp. Using two colonoscopic frame datasets, we demonstrate the experimental results that show the significant reduction in the number of false positives by adding selected regions in our computer-aided polyp segmentation system.
|
| 12 |
,Image GPT with Super Resolution, |
Bhumika Shah,Ankita Sinha,Prashant Saxena |
|
Abstract
A Generative Pre-trained Transformer (GPT) model which can generate text by looking at previous text was trained to generate image pixels sequentially by making a correlation between the image classification accuracy and the image quality. This model uses the generative model for generating images. The Image Generative Pre-trained Transformer (IGPT) works on a low-resolution image which in turn produces a low-resolution output. In this paper, we have attempted to eliminate this limitation by enhancing the resolution of the output image produced by IGPT. The primary focus during this research work is to check different models and choose the simplest model for improving quality of the image generated because there are several models that support deep neural networks that have been successful in upscaling the image quality with great accuracy for achieving super resolution for a single image. The output image of low resolution is upscaled to high-resolution space employing a single filter and bicubic interpolation. We have also considered peak signal-to-noise ratio (PSNR) score and structural similarity (SSIM) value to analyze the standard of the image produced by the algorithm. The p
|
| 13 |
,Boosting Accuracy of Machine Learning Classifiers for Heart Disease Forecasting, |
Divya Lalita Sri Jalligampala,R. V. S. Lalitha,M. Anil Kumar,Nalla Akhila,Sujana Challapalli,P. N. S |
|
Abstract
Heart disease is one of the significant diseases which causes a huge number of deaths all over the world. Even medical specialists are facing difficulties for the proper diagnosis of heart disease which raises a need for a new classification scheme. But it becomes a crucial task for healthcare providers due to the rapid increase of medical data size every day. To resolve this, several machine learning algorithms are discussed in this paper, and these algorithms’ performance is measured by using different metrics like accuracy, precision, recall, and F1-score. But these algorithms are not acceptable for accurate prediction and diagnosis. To further improve the accuracy of classifiers, different ensemble methods were used because for any machine learning algorithm, accuracy is the main criteria to measure the performance. In this new methodology, the feature importance method is used as a pre-processing technique to get a minimum number of attributes rather than using all attributes in the dataset which has impact on the accuracy of classifiers. After that pre-processed data is trained by using various classifiers like linear regression, SVM, naïve Bayes, and decision tree, and then
|
| 14 |
,Rapid Detection of Fragile X Syndrome: A Gateway Towards Modern Algorithmic Approach, |
Soumya Biswas,Oindrila Das,Divyajyoti Panda,Satya Ranjan Dash |
|
Abstract
Human medical care is perhaps the main subjects for society. It attempts to track down the right powerful and strong infection location straightaway to patients in the proper considerations. With an anticipated birth prevalence of 0.12 in 1000 females and 0.25 in 1000 males, the second most frequent cause of serious mental disability is Fragile X Syndrome. It’s a repetition of trinucleotide, when (CGG)n factor located within the 5’ untranslated region of the Fragile X Mental Retardation 1 (FMR1) gene develops to more than 200 repetitions (complete mutation) and becomes hypermethylated. The failure to transport FMR1 protein (FMRP), which occurs in the fragile X condition, is linked to such events. Since the wet lab technique is not precised enough and time consuming, so the dry lab methods like statistics, bioinformatics and computer science are becoming fundamental gateway for disease diagnosis and new model towards modern treatment. Here, we have proposed a gene alignment and nucleotide base matching algorithm ‘FXSDetect’ managing the determination of Fragile X syndrome in very short span leading a way towards rapid diagnosis.
|
| 15 |
,Summarizing Bengali Text: An Extractive Approach, |
Satya Ranjan Dash,Pubali Guha,Debasish Kumar Mallick,Shantipriya Parida |
|
Abstract
Text summarization is a challenging task in the field of Natural Language Processing. In the case of lower resource language like Bengali is also a difficult task to make an automatic text summarization system. This paper is based on the extractive summarization of Bengali text. Text summarization deletes the less useful information in a piece of text or a paragraph and summarizes it into a confined text. This helps in finding the required text more effectively and quickly. There are many types of algorithms used for summarizing the text. Here in this paper, we have used TF-IDF and BERT-SUM technologies for the Bengali extractive text summarization.
|
| 16 |
,Dynamic Hand Gesture Recognition of the Days of a Week in Indian Sign Language Using Low-Cost Depth |
Soumi Paul,Madhuram Jajoo,Abhijeet Raj,Ayatullah Faruk Mollah,Mita Nasipuri,Subhadip Basu |
|
Abstract
We develop a dynamic hand gesture recognition system on the gestures of seven days of the week in Indian Sign Language. We use Kinect V2 sensor, which is a low-cost RGB-D camera, to collect videos of gestures of two subjects and then perform key frame extraction, cropping of static regions, background subtraction, feature extraction and classification. On the resultant dataset, tenfold cross-validation by random forest classifier gives us an accuracy of around 74.29%.
|
| 17 |
,Sentiment Analysis on Telugu–English Code-Mixed Data, |
|
|
Abstract
|
| 18 |
,Fuzziness on Interconnection Networks Under Ratio Labelling, |
A. Amutha,R. Mathu Pritha |
|
Abstract
Investigation of interconnection networks like circulant network, hypercubes .(.), cube connected cycles CCC(.), butterfly network BF(.), beans networks for the admissibility fuzziness is the notion of the paper. The binary tree, star graph do not admit fuzziness under ratio labelling. Classification of these interconnection networks as Cayley graph leads to a conclusion that not all Cayley graphs are fuzzy graphs under ratio labelling.
|
| 19 |
,CoviNet: Role of Convolution Neural Networks (CNN) for an Efficient Diagnosis of COVID-19, |
D. N. V. S. L. S. Indira,R. Abinaya |
|
Abstract
Coronaviruses are a large family of viruses that can cause a human being to become critically sick. In different forms, COVID-19 affects various people. Since COVID-19 has begun its rampant expansion, isolating COVID-19 infected individuals is the best way to deal with it. This can be accomplished by monitoring individuals by running recurring COVID tests. The use of computed tomography (CT-scan) has demonstrated good results in evaluating patients with possible COVID-19 infection. Patients with COVID will heal with the help of antibiotic therapy from vitamin C supplements. Patients with these symptoms need a faster response using non-clinical methods such as machine learning and deep neural networks in order to manage and address additional COVID-19 spreads worldwide. Here, in this paper we are diagnosis the covid-19 patients with CT-scan images by applying XGBoost classifier. Developed a web application which basically accepts a patient CT-scan to classify COVID positive or negative. After that, the negative class patients with symptoms are suggested with a danger rate with the help of age groups, health-related issues, and the area he/she belongs to. Three machine learning algor
|
| 20 |
,Deep Learning for Real-Time Diagnosis of Pest and Diseases on Crops, |
Jinendra Gambhir,Naveen Patel,Shrinivas Patil,Prathamesh Takale,Archana Chougule,Chandra Shekhar Pra |
|
Abstract
Agriculture is the main stay of India as it contributes significantly to the economy by providing employment to more than half of country’s workforce. The major problem faced by farmers during the crop production is the pest and disease attack. Lack of right technical advice at right time leads to improper farm decision leading to economic losses. It is necessary to provide an interface to farmers to identify the pest and disease problem faced during agricultural processes and get a solution to that problem from agricultural experts. We have developed an interface in the form of an Android application to upload the images and a Web interface to display the uploaded images with their disease or pest type. The uploaded images are used to train a CNN model which will help to identify pests and diseases for the crops. This system is currently being developed for four major crops, which are rice, maize, chickpea, and lentil.
|
|
|
发表于 2025-3-21 18:53:35
