K

KYC Document Verification With Live Video

The KYC Verification System achieves the following objectives - 1) Perform face matching using live video and government ID proof 2) To check for authentication of the ID proof provided

K

KYC Document Verification With Live Video

The KYC Verification System achieves the following objectives - 1) Perform face matching using live video and government ID proof 2) To check for authentication of the ID proof provided

The problem KYC Document Verification With Live Video solves

We are moving towards completely digital onboarding of customers across various sectors. Most businesses such as banks need one or more of the customer documents to identify the authenticity of the customer. There are continued fraud reports about customers uploading documents of some third party, deceased individuals, etc. without their consent leading to forged accounts being opened. This problem can be resolved by digitally scanning all the documents (which are ways to validate the authenticity) such as PAN, Driving License, Passport, AADHAR, etc, extracting customer photos from the document and comparing them with images from a live 30/20-sec video. This will ensure that documents belong to the person uploading them.
The project will reduce the manpower required for KYC Verification process by the business and in turn provide a more faster and secured way of completing a user/customer's KYC.
Moreover the user can also complete the process of KYC in a quicker and safer way by sitting at the comfort of their home and completing the KYC process on their suitable time. It will also relieve the customer from going to physical centres for KYC Verification in these pandemic times wherein they can complete the process online.

Challenges we ran into

They were many challenges faced during the building of this project

  1. Coming up with a simple design and formulating simple steps for user to do their KYC verification successfully
  2. Market Survey of the existing products similar to our project and Literature survey for choosing the best algorithm i.e. DeepFace algorithm suitable to our project for face matching and face verification
  3. Fixing the bug for extracting data from the QR code given in the Adhaar card of the user and verifying it with details provided by the user to authenticate the ID proof and adding one more layer of verification of the user.

Discussion