Search

This work is an attempt to solve one of the most common problems of scene recognition. The dataset called PLACES2 was used for the purpose of experimentation which contains nearly 7 million labelled pictures of various scenes which is a sufficiently exhaustive dataset to train and test a model to get the desired output. The Convolutional Neural Network was used to design a neural architecture for our machine to store and classify the images as desired for the recognition of scene in the dataset. Results strongly suggest that the trained model were able to perform successful scene recognition in the given data sample

A potential rise in interest in the Internet of Things in the upcoming years is expected in the fields of healthcare, supply chain, logistics, industries, smart cities, smart homes, cyber physical systems, etc. This paper discloses the fusion of the Internet of Things (IoT) with the so-called “distributed ledger technology” (DLT). IoT sensors like temperature sensors, motion sensors, GPS or connected devices convey the activity of the environment. Sensor information acquired by such IoT devices are then stored in a blockchain. Data on a blockchain remains immutable however its scalability still remains a challenging issue and thus represents a hindrance for its mass adoption in the IoT. Here a communication system based on IOTA and DLT is discussed with a systematic architecture for IoT devices and a future machine-to-machine (M2M) economy. The data communication between IoT devices is analyzed using multiple use cases such as sending DHT-11 sensor data to the IOTA tangle. The value communication is analyzed using a novel “micro-payment enabled over the top” (MP-OTT) streaming platform that is based on the “pay-as-you-go” and “consumption based” models to showcase IOTA value transactions. In this paper, we propose an enhancement to the classical “masked authenticated message” (MAM) communication protocol and two architectures called dual signature masked authenticated message (DSMAM) and index-based address value transaction (IBAVT). Further, we provided an empirical analysis and discussion of the proposed techniques. The implemented solution provides better address management with secured sharing and communication of IoT data, complete access control over the ownership of data and high scalability in terms of number of transactions that can be handled.

Twitter is extensively used as an information-sharing platform during any kind of emergency like disasters etc. People tweet useful information about disaster-related events such as evacuations, volunteer need, help, warnings etc. This data is sometimes very useful for rescue teams, NGOs, military and various other government and private organisations who are tasked with responsibilities to save lives and provide volunteers. This data can also be used to analyze disaster behaviour. In this paper, we have collected labelled tweets from crisisLexT26 and crisisNLP and classified them into seven labels on the basis of information provided by them. The data was heavily skewed. So to improve the accuracy of classifiers, we have applied various techniques as a result of which we have created two datasets (Imbalanced and Balanced). We have compared the performance of various BERT-based models on these two datasets. For sequence classification, a balanced dataset performs better than an imbalanced dataset. We can improve accuracy of classifiers to great extent by adopting good data preprocessing and data splitting techniques.

New age advancement seen in the ICT has enthralled human enthusiasm and excitement to the next level of getting informed and educated about the things they interact with. We want to capture the history of the things from their point of origin. But it has been that history can be altered and misinterpreted for self means. Blockchain is one such revolutionary architecture that ensures history is preserved by maintaining concoction of consortium or consensus over the network between all nodes for historical data and ensures immutability. Different blockchain architecture gives different types of flexibility and usability. Traceability is must for accountability for ensuring history is preserved. Supply chain for pharmaceutical drugs is one such problem where customers need to know the exact timeline of the medicine they are about to intake. Counterfeit drugs are big issue in the market and it is bigger problem than we realize. Fake drugs are entering into the mainstream supplies in the country as traceability is never assured from any other technology. This chapter solves this problem by jotting down each transaction into an immutable distributed ledger as blockchain architecture which cannot be altered or changed. Whenever there is a change of hand of medicines, two things are being recorded, first represent the current owner, and latter remain the timestamp when this took place. Ethereum and Hyperledger Fabric along with Hyperledger Composer are the two building blocks of the Blockchain platform used in this chapter. This chapter shows the comparative difference between the two under the use case of pharmaceutical supply chain and defines what the adoptive approach should be while choosing between the two platforms. This chapter shares some pseudo codes for the two platforms showing how to interact with these platforms. At the end, some challenges faced are discussed along with future scope and direction to research areas with some additional resources listed as references.

Background Consumer electronics or daily use home appliances are the basic necessity of every household. With the adoption of IoT in consumer electronics, this industry is set to rise exponentially. In recent times, the demand for consumer electronics rises amidst the pandemic due to a paradigm shift from in-office culture to work from home. Despite intelligent IoT devices, smart home configuration, and appliances at our disposal, the rudimentary client-server architecture fails to provide facilities like full access control of data and devices, transparency, secured communication, and synchronization between multiple devices, etc. to the users. Methods To overcome these limitations, Blockchain technology has been adopted in recent years, however, it has its own set of limitations in its widespread implementation. Hence, we propose a methodology using the IOTA platform, a distributed ledger technology (DLT) for secured communication between consumer electronics devices and appliances. Results The implementation provides access control, interoperability, data storage, and management with an exploratory insight towards a decentralized micro-payment use-case between Electric cars and charging stations.

INTRODUCTION: Parkinson's disease (PD) occurs due to the deficiency of dopamine that regulates various activities ofthe human body. Researchers have identified that voice is an underlying symptom of PD. Recently, Machine learning(ML) has helped in solving problems of computer vision, natural language processing, speech recognition etc.OBJECTIVES: This paper aims to analyse the effect of feature type selection i.e. MFCC and TQWT on the efficiency ofvoice based PD detection system along with the use an ensemble learning based classifier for this task.METHODS: Hence, in this work, various machine learning models, including Logistic Regression, Naive Bayes, KNN,Random Forest, Decision Tree, SVM, MLP, and XGBoost, have been employed and explored for PD detection purpose.The task of Feature selection was also done using minimum-Redundancy and Maximum-Relevance (mRMR) andRecursive Feature Elimination (RFE) techniques.RESULTS: The results of the XGBoost with mRMR feature selection, outperformed all other models with a high accuracyof 95.39% and precision, recall and F1-score of 0.95 each, when both MFCC and TQWT features were selected.CONCLUSION: The results obtained strongly support the use of XGBoost model for the voice sample based detection ofPD along with mRMR feature selection technique.

Isolated Word Recognition is the process of converting the spoken word into its corresponding text format. At present mainly Mel Frequency Cepstrum Coefficients (MFCC) is used as the feature extraction parameter i.e. the identifying features for the speech signal. Through this paper efforts have been made to determine the accuracy of an MFCC based system and also to build an isolated word recognizer based on word acoustic model that uses MFCC in combination with other features of speech such as Root Mean Square Energy, Length of the word and its Brightness. Using an artificial neural network as the classifier, the system was trained & tested for a set of spoken isolated words. The results obtained showed a high and an increased accuracy for the experiment in which along with MFCC other selected parameters were also involved against the experiment which only involved MFCC.

Substitution-boxes are the only source of nonlinearity in various symmetric encryption systems and responsible for inducing confusion of plaintext data. The robustness of these systems exclusively depends on the potentiality of S-boxes. The design methods having fast and simple computations which can yield effective S-boxes are preferred. In this paper, a new chaos-based random search is applied to construct cryptographically potent 8 × 8 S-box. The method explores the features of piecewise linear chaotic map for candidate generation and random search. The optimized S-box obtained is tested against standard statistical tests like bijectivity, equiprobable I/O XOR distribution, nonlinearity, and strict avalanche criteria revealing its superior performance. The proffered substitution-box is further compared with some contemporary chaotic substitution-boxes. The results confirm a consistent design, suitable for building strong block encryption systems.