Audio Tagging Using CNN Based Audio Neural Networks for Massive Data Processing

Sound event detection, speech emotion classification, music classification, acoustic scene classification, audio tagging and several other audio pattern recognition applications are largely dependent on the growing machine learning technology. The audio pattern recognition issues are also addressed by neural networks in recent days. The existing systems operate within limited durations on specific datasets. Pretrained systems with large datasets in natural language processing and computer vision applications over the recent years perform well in several tasks. However, audio pattern recognition research with large-scale datasets is limited in the current scenario. In this paper, a large-scale audio dataset is used for training a pre-trained audio neural network. Several audio related tasks are performed by transferring this audio neural network. Several convolution neural networks are used for modeling the proposed audio neural network. The computational complexity and performance of this system are analyzed. The waveform and leg-mel spectrogram are used as input features in this architecture. During audio tagging, the proposed system outperforms the existing systems with a mean average of 0.45. The performance of the proposed model is demonstrated by applying the audio neural network to five specific audio pattern recognition tasks.

Early Stage Detection of Crack in Glasses by Hybrid CNN Transformation Approach

Recently, glass crack detection methods have been emerging in Artificial intelligence programming. The early detection of the crack in glass could save many lives. Glass fractures can be detected automatically using machine vision. However, this has not been extensively researched. As a result, a detection algorithm is a benefit to study the mechanics of glass cracking. To test the algorithm, benchmark data are used and analysed. According to the first findings, the algorithm is capable of figuring out the screen more or less correctly and identifying the main fracture structures with sufficient efficiency required for majority of the applications. This research article has addressed the early detection of glass cracks by using edge detection, which delivers excellent accuracy in fracture identification. Following the pre-processing stage, the CNN technique extracts additional characteristics from the input pictures that have been provided due to dense feature extraction. The "Adam" optimizer is used to update the bias weights of networks in a cost-effective manner. Early identification is achievable with high accuracy metrics when using these approaches, as shown in the findings and discussion part of this paper.

Automatic Car Damage detection by Hybrid Deep Learning Multi Label Classification

Automating image-based automobile insurance claims processing is a significant opportunity. In this research work, car damage categorization that is aided by the hybrid convolutional neural network approach is addressed and hence the deep learning-based strategies are applied. Insurance firms may leverage this paper's design and implementation of an automobile damage classification/detection pipeline to streamline car insurance claim policy. Using deep convolutional networks to detect car damage is now possible because of recent improvements in the artificial intelligence sector, mainly due to less computation time and higher accuracy with a hybrid transformation deep learning algorithm. In this paper, multiclass classification proposed to categorize the car damage parts such as broken headlight/taillight, glass fragments, damaged bonnet etc. are compiled into the proposed dataset. This model has been pre-trained on a wide-ranging and benchmark dataset due to the dataset's limited size to minimize overfitting and to understand more common properties of the dataset. To increase the overall proposed model’s performance, the CNN feature extraction model is trained with Resnet architecture with the coco car damage detection datasets and reaches a higher accuracy of 90.82%, which is much better than the previous findings on the comparable test sets.

Vehicle Navigation System based on Pollution Metric Analysis with Q-Learning Algorithm

The navigation systems available in the present scenario takes into account the path distance for their estimations. In some advanced navigation systems, the road traffic analysis is also considered in the algorithm for their predictions. The proposed work estimates a navigation path with respect to the present pollution level on the roadways. The work suggests an alternate path to avoid additional vehicles to enter the same road which is already impacted by air pollution. A Q-learning (Quality learning) prediction algorithm is trained in the proposed work with a self-made dataset for the estimations. The experimental work presented in the paper explores the accuracy and computational speed of the developed algorithm in comparison to the traditional algorithms.

An Efficient Machine Learning based Model for Classification of Wearable Clothing

Computer vision research and its applications in the fashion industry have grown popular due to the rapid growth of information technology. Fashion detection is increasingly popular because most fashion goods need detection before they could be worn. Early detection of the human body component of the input picture is necessary to determine where the garment area is and then synthesize it. For this reason, detection is the starting point for most of the in-depth research. The cloth detection of landmarks is retrieved through many feature items that emphasis on fashionate things. The feature extraction can be done for better accuracy, pose and scale transmission. These convolution filters extract the features through many epochs and max-pooling layers in the neural networks. The optimized classification has been done using SVM in this study, for attaining overall high efficiency. This proposed CNN approach fashionate things prediction is combined with SVM for better classification. Furthermore, the classification error is minimized through the evaluation procedure for obtaining better accuracy. Finally, this research work has attained good accuracy and other performance metrics than the different traditional approaches. The benchmark datasets, current methodologies, and performance comparisons are all reorganized for each piece.

Smart Medical Nursing Care Unit based on Internet of Things for Emergency Healthcare

Recently, the development and integration of various sensor control with smart intelligent unit is used in medical field through IoT. However, there is still a lot of space for growth in the medical and health industry's use of new technology. The traditional nurse care unit is managed through medical staffs, and the expanding medical demands creates the hospital’s patients records to be updated inefficiently. Since this is now an urgent need, developing a realistic, smart medical nursing care unit at low cost with a system capable of facilitating the effective and convenient administration of medical staff has taken a new significance. The proposed framework, conducted in the analysis to monitor medical records and activities of the emergency care unit patients, functions as a nurse and gives patients the nurse care satisfaction. The patients' actual location may be obtained for the first time by cloud computing based smart system. The precise location of the patient is critical to rescue the patient in emergency situation. This research work illustrates that the intelligent nurse care unit is the main phase called Smart Medical Nursing Care (SMNC). It contains several sensor units and by the combination of many sensors in the sensor module, it takes very less reaction time to connect or communicate both sides i.e., between patients and medical staffs.

Smart Fashion: A Review of AI Applications in Virtual Try-On & Fashion Synthesis

The rapid progress of computer vision, machine learning, and artificial intelligence combined with the current growing urge for online shopping systems opened an excellent opportunity for the fashion industry. As a result, many studies worldwide are dedicated to modern fashion-related applications such as virtual try-on and fashion synthesis. However, the accelerated evolution speed of the field makes it hard to track these many research branches in a structured framework. This paper presents an overview of the matter, categorizing 110 relevant articles into multiple sub-categories and varieties of these tasks. An easy-to-use yet informative tabular format is used for this purpose. Such hierarchical application-based multi-label classification of studies increases the visibility of current research, promotes the field, provides research directions, and facilitates access to related studies.

Probabilistic Neural Network based Managing Algorithm for Building Automation System

A building automation system is a centralized intelligent system, which controls the operation of energy, security, water, and safety by the help of hardware and software modules. The general software modules employed for automation process have an algorithm with pre-determined decisions. However, such pre-determined decision algorithms won’t work in a proper manner at all situations like a human brain. Therefore a human biological inspired algorithms are developed in recent days and termed as neural network algorithms. The Probabilistic Neural Network (PNN) is a kind of artificial neural network algorithm which has the ability to take decisions same as like of human brains in an efficient way. Hence a building automation system is proposed in the work based on PNN for verifying the effectiveness of neural network algorithms over the traditional pre-determined decision making algorithms. The experimental work is further extended to verify the performances of the basic neural network algorithm called Convolution Neural Network (CNN).

Deniable Authentication Encryption for Privacy Protection using Blockchain

Cloud applications that work on medical data using blockchain is used by managers and doctors in order to get the image data that is shared between various healthcare institutions. To ensure workability and privacy of the image data, it is important to verify the authenticity of the data, retrieve cypher data and encrypt plain image data. An effective methodology to encrypt the data is the use of a public key authenticated encryption methodology which ensures workability and privacy of the data. But, there are a number of such methodologies available that have been formulated previously. However, the drawback with those methodologies is their inadequacy in protecting the privacy of the data. In order to overcome these disadvantages, we propose a searchable encryption algorithm that can be used for sharing blockchain- based medical image data. This methodology provides traceability, unforgettable and non-tampered image data using blockhain technology, overcoming the drawbacks of blockchain such as computing power and storage. The proposed work will also sustain keyword guessing attacks apart from verification of authenticity and privacy protection of the image data. Taking these factors into consideration, it is determine that there is much work involved in providing stronger security and protecting privacy of data senders. The proposed methodology also meets the requirement of indistinguishability of trapdoor and ciphertext. The highlights of the proposed work are its capability in improving the performance of the system in terms of security and privacy protection.