The problem Espial solves

Surveillance ensures safety and security, an integral part of every individual’s life, That's why cameras are installed everywhere. But still news of crime resonates throughout the year. Who is looking at those footage prior to an incident? Are they just an evidence collecting source? To address these issues, we present to you our project: Espial.
The motive behind our idea is to prevent crimes by sending real time alerts to the users, in residential areas, public places etc. A deep learning model integrated with your application, and personalised user login which detects and sends the notifications in real time to the concerned authority in case any suspicious activity is detected is the main highlight of this project.

• Our model uses Computer Vision to narrow down to the target for faster and more reliable predictions.
  We used OpenCV, a Computer Vision library, to track the movement of people in real time using Multiple Object Detection through Localisation.
• Deep Learning models like Customised CNN Model, VGG Net, AlexNet, Inception+CustomisedCNN were trained on a dataset which consisted of around 2200 videos. The dataset we used was seeded from combined datasets scrapped from different sources, comprising videos taken from various angles, enclosing all types of violent and non-violent natured clips.
• If any suspicious activity is found, the result is sent to the user as a notification along with a short recorded video of the captured moment.

Challenges we ran into

• **Adding Support for multiple cameras : ** To make our project more robust and application-oriented, multiple camera support was a key feature. However due to restrictions in processing power and notification management using web, it proved to be a very difficult task. A unique token based authentication system was thus used in our project to authenticate and verify the cameras.
• **Finding the Dataset : ** Suspicious activities are a very subjective matter, and thus finding a good dataset and training a model which can actually detect them was a crucial and tough task in this project.