ISSN 2348-1196 (print) International Journal of Computer Science and Information Technology Research ISSN 2348-120X (online) Vol. 8, Issue 1, pp: (47-53), Month: January - March 2020, Available at: www.researchpublish.com
SECURITY ANALYSIS IN CONVOLUTIONAL NEURAL NETWORK USING TIME SERIES CLASSIFICATION AND CRYPTO SCHEMES 1
S.KAVITHA, 2A.SENTHILKUMAR
1
Research scholar, Dept. of. Computer science, Tamil University, Thanjavur-613010.
2
Asst.professor, Dept. of .Computer Science,Tamil University (Established by the Govt.of.Tamilnadu),Thanjavur-613010
Abstract: For image based analysis in network implications, an important verified visual representation are applied which are termed as, ‘Convolution Neural Networks’. The existing Convolution Neural Networks process involves , ‘Time Series Classification’ were an image depicted as the data set are predicted to extract a pattern. Based on this pattern, exact specification to predict a concrete solution are allowed in the existing schemes. To enhance the existing system and to propose novelty in adapting the information security process. The existing system is adapted to undergo the process of appending secrecy bits in the original image are executed in this work. Further the secret pattern orientated image are transmitted to all other participating nodes, which ensures a secured transmission in the networks. After the quantization analysis of the existing models the secrecy image representation has produced considerably significant percentage of 5% increased protection measures with respect to its dataset image representation. Hence the proposed system evaluates the time series classification that are commonly used in Convolution Neural Networks and adopts security enhancement which proves that data are transmitted in a secured manner by utilizing the cryptography methods. Keywords: Neural Network,, Cryptography , Information security .
1. INTRODUCTION Convolutional Neural Networks (CNNs) have been established as a robust class of models for image recognition problems. Encouraged by these results, we provide an extensive empirical evaluation of CNNs on large-scale video classification using a new dataset of 1 million YouTube videos belonging to 487 classes. We study multiple approaches for extending the connectivity of a CNN in time domain to take advantage of local spatio-temporal information and suggest a multiresolution, forested architecture as a promising way of speeding up the training. Our best spatio-temporal networks display significant performance improvements compared to strong feature-based baselines (55.3% to 63.9%), but only a surprisingly modest improvement compared to single-frame models (59.3% to 60.9%). We further study the generalization performance of our best model by retraining the top layers on the Action Recognition dataset and observe significant performance improvements compared to the UCF-101 baseline model. The system introduces a class of efficient models called Mobile Nets for mobile and embedded vision applications. Mobile Nets are based on a streamlined architecture that uses depth-wise separable convolutions to build light weight in deep neural networks. We introduce two simple global hyper-parameters that efficiently trade off between latency and accuracy. These hyper-parameters allow the model builder to choose the right sized model for their application based on the constraints of the problem. We present extensive experiments on resource and accuracy tradeoffs and show strong performance compared to other popular models on Image Net classification. The effectiveness of Mobile Nets known across a wide range of applications are considered as an example and use cases including object detection, fine grain classification, face attributes and large scale geo-localization. Rendering the semantic content of an image in different styles is a difficult image processing task.
Page | 47 Research Publish Journals