Face Recognition Using Deep Learning

Security Banking Industry


The use case here is to build a realtime Solution using AI which will aid in detection of crimes and malicious activities using face recognition patterns.

image



Summary

  • Our client is from the banking domain who are majorly looking for a solution for their banking Services. Their solution involves advanced usage of Artificial intelligence along with computer vision to make sure that the detection of counterfeiting activities are accurate and efficient.
  • The client was specific about their requirement on the accuracy of results and also about the real time capabailties of the solution that was being built.
  • Our team designed and implemented the solution using six plus technologies which included Machine Vision, Micro Radar, Signal Processing, Deep Learning and Artificial Intelligence.

Tech Challenge

  • One of the major challenges with the solution was multiple human detection in the surrounding area.
  • Another challenge of this use case was tampering the camera video which could result in failure of the solution.
  • Giving real time alerts to so that such counterfeiting activities can be avoided.
  • Using multiple complex technologies to get to the desired accuracy.

Solution

  • The system is capable enough to give real time alerts to the admin to take necessary action.
  • Human Detection and Collision was managed using Computer Vision and AI Algorithms.
  • The solution uses the time tracking algorithm to detect regular image recognitions vs unwanted activities.
  • The system is automated to give process signals and data and give alerts.
  • Metrics gathered are in real time and are very detailed and specific which included dimensions of human body, time taken, voice in decibel etc.

Project Planning


  • 2 Week -Solution Architecture Desgin - Solution Architect

  • 4 Week - Integration, Testing and Deployment

  • 6 Week - Exploratory Data Analysis and Model Building

  • 8 Week - Collection of data and realtime streamlining and development

Results

  • We have developed a solution that is backed with a lot of advanced technologies and deep learning algorithms thus making the solution accurate as well as efficient.
  • The solution is capable of processing large amounts of data coming from different resources.
  • The solution allowed the customers to get real time alerts using all possible ways of detecting a theft or malicious activity .

Send Us A Note

Have an idea? Let's TALK