Video Steganography Using Convolution Neural Networks and Temporal Residual Modelling

Abstract

Today, one of the top priorities of governments is traffic safety. Given the importance of the subject, identifying the causes of road accidents has become the primary goal in reducing the damage caused by traffic accidents. Machine learning and data mining concepts are used to recognize the various factors that influence road accidents and their severity. The application uses at odd Machine Learning algorithms which includes K-Nearest-Neighbor, Decision tree, Random forest, etc make predictions based on various parameters, within all these models KNN gives the best accuracy. This information can be used to analyze future inputs and improve the system's output accuracy. This model can be improved further to send the accident report to the appropriate authorities, such as hospitals, ambulances, and insurance companies, and can thus be very useful in reducing accident fatality rates in the country.