Vibration Guided Automatic Vision for Enhanced Security

Abstract

The existing security systems are secure but are not smart enough to handle arbitrary scenarios leading to many false triggers of the alert system. Furthermore, these systems require constant human intervention which isdifficult to achieve.They are also vulnerable as they contain many loopholesand the sensors used are easily manipulatable. The proposed system tries to solve this problem in an efficient and a smart way by the use of sensors, AI and IoT which makes the system robust and resistant againstattacks. The system implements advanced face detection via Single Shot Detection and face recognition via Inception Neural Network for recognition of object in a fast and accurate way. This helps the system act according to the situation, thus preventing any damage to theregion which implements this system. In this work the proposed system is implemented and tested as a Home Security System. The system can also be extended to work in other areas like banks, data hubs, museums etc.The overall accuracy of the system was recorded to be 97.95%.