Background: In making the deep neural network, activation functions play an important role. But the choice of activation functions also affects the network in term of optimization and to retrieve the better results. Several activation functions have been introduced in machine learning for many practical applications. But which activation function should use at hidden layer of deep neural networks was not identified. Objective: The primary objective of this analysis was to describe which activation function must be used at hidden layers for deep neural networks to solve complex non-linear problems. Methods: The configuration for this comparative model was used by using the datasets of 2 classes (Cat/Dog). The number of Convolutional layer used in this network was 3 and the pooling layer was also introduced after each layer of CNN layer. The total of the dataset was divided into the two parts. The first 8000 images were mainly used for training the network and the next 2000 images were used for testing the network. Results: The experimental comparison was done by analyzing the network by taking different activation functions on each layer of CNN network. The validation error and accuracy on Cat/Dog dataset were analyzed using activation functions (ReLU, Tanh, Selu, PRelu, Elu) at number of hidden layers. Overall the Relu gave best performance with the validation loss at 25th Epoch 0.3912 and validation accuracy at 25th Epoch 0.8320. Conclusion: It is found that a CNN model with ReLU hidden layers (3 hidden layers here) gives best results and improve overall performance better in term of accuracy and speed. These advantages of ReLU in CNN at number of hidden layers are helpful to effectively and fast retrieval of images from the databases.

In last few years' terrorist and illegal attacks across many country borders has increased which led to security and strict passport verification process. This turned down legitimate travelers. Many countries are in process of implementing electronic passports to travelers for travelling, automating passport verification process and increasing border security. The e-passport deploys two popular technologies: Radio frequency Identification (RFID) and Biometrics. Personal credentials and bearers biometric data is stored on RFID chip which is used in verification process by border security officers. The next generation of e-passports will implement more advanced cryptographic mechanisms, collectively known as Extended Access Control, and in particular a protocol referred to as Chip Authentication that protects an e-passport against cloning and transferability attacks. The Extended Access Control suite of protocols has found minor attention in the literature until now. The paper analyses the study of various technologies used in Epassport design. A cryptographic security analysis of the epassport using face fingerprint, palm print and iris biometric that are intended to provide improved security in protecting biometric information of the e-passport bearer. Together, RFID and biometric technologies promise to reduce fraud, ease identity checks, and enhance security. At the same time, these technologies raise new risks. We explore the privacy and security implications of this worldwide implementing next-generation authentication technology: e-passport. We describe privacy and security issues that apply to e-passports, and then analyze these issues in the context of the International Civil Aviation Organization (ICAO) standard for e-passports.

In the recent years, the development in computer technologies and multimedia applications has led to the production of huge digital images and large image databases, and it is increasing rapidly. There are several different areas in which image retrieval plays a crucial role like Medical systems, Forensic Labs, Tourism Promotion, etc. Thus retrieval of similar images is a challenge. To tackle this rapid growth in digital repositories it is necessary to develop image retrieval systems, which can operate on large databases. There are basically three techniques, which is useful for efficient retrieval of images. With these techniques, the number of methods has been modified for the efficient image retrieval of images. In this paper, we presented the survey of different techniques that has been used starting from Image retrieval using visual features and latest by the deep learning with CNN that contains the number of layers and now becomes the best base method for retrieval of images from the large databases. In the last section, we have made the analysis between various developed techniques and showed the advantages and disadvantages of various techniques.

In the concepts of Deep Learning, the ConvNets are a classified group of deep neural networks that are used to analyze the images. They find their major applications in image classification and in the medical image analysis. ConvNets are regularized, multilayer preceptors. The major challenge today is the increasing size of medical image repositories which is leading to troubles in large database management and then querying those databases for storage and retrieval and this can only be sorted out using the Content Based Medical Image Retrieval (CBMIR) systems. A major challenge in CBMIR systems is the semantic gap that exists between the low-level visual information captured by the imaging devices and high-level semantic information captured by human. This paper demonstrates a new and efficient framework with the algorithm of Deep CNN for the Feature Based Medical Image Retrieval (FBMIR) system for fast and efficient retrieval of medical and clinical images for detecting Pneumonia. The intermodal datasets are used and divided into two classes to train the network. One class consists of the lungs infected by Pneumonia and the other class contains the images of lungs which are normal. The learned features from the trained deep convolutional neural network with the fast processing feature of the hash method are used to reduce the feature space. To achieve the best retrieval results with class-based predictions 93.9% of average classification accuracy is achieved and 0.87 of mean average precision is achieved for this retrieval task. The proposed framework fit the best for top-ranked multimodal medical image retrieval.