Real-time Transformer Vandalism Detection by Application of Tuned Hyper Parameter Deep Learning Model

Vandalism is an illegal act of cannibalism or change face to a private public property by human beings for re-sale parts punish the owner. Initial research findings on transformer detection have fallen short image recognition vandal in real-time but only does activities after damage done as it occurs. Automated systems using sensor feed trained deep learning model new vandalism approach with capabilities three-dimensional learning, extracting important features automatically and temporal output prediction. This paper aims at distinguishing object entering zoned area without permission take away modify established infrastructure, so that Vandal can be arrested before causing any transformer. The researchers identified multiplicative hybrid combining convolutional neural networks long short-term memory application problem detect enters restricted installation site. accuracy improved tuning hyper-parameters specific considered this work are number layers epochs. applying taken Image model. method increases prediction from input data lowers computational processing complications due reduced volume through pooling. system validated ImageNet dataset. Results achieved five sixty epochs 99% accuracy. performance increased respectively was best result compared lower Further increase these parameters resulted overfitting.