The project follows a methodology of creating a system that is cost-effective and has fewer hardware requirements. Thus, without making changes to the existing system this system can be used for ATM surveillance. The main challenge the team faced was the availability of data. Since, the idea was to make the system capable of real-time video processing, at every phase of design and development, the overall cost was kept low so that the system becomes effective. One limitation was the hardware requirement(availability of GPUs) for training purposes. We also have problems integrating real-time streaming with our backend model.