Face Recognition Using Big Data Suite

Technology Trends & Markets

Biometrics, Bigdata Analytics, MySQL, MongoDB, Casandra Neo4J, Paraccel, Amazon web services, Rackspace, Open stack, Tablue, cloudera, Java, Python, PHP, R.
Markets: Retail, Media, Healthcare, Law enforcement, Education, Banking, Telecom, Government, Manufacture, transportation.

Our Products

1. IFRI (Insights of Data Face Recognition Identification)

IFRI is still a image based recognition system. This application can now be deployed from a centrally managed platform, resulting in fast and efficient identity resolution.

Comprehensive Facial Recognition Solution
 Technology transforms lat 2D facial images into 3D avatar for better results.
 Powerful search and analysis capabilities match them against the enrolled images in the IFRI database or third-party database.
 Provides all management and control for Insights of data IFRI applications.

Features:
 Advanced 3D facial recognition for smartphones and web connected digital devices.
 Real-time in field ID processing.
 Real-time watch list search.
 Advanced face biometrics for uncontrolled and low resolution facial Imagery.
 Forensic analysis and comparative tools to analyze and match single/multiple facial images in 2D/3D modes.

USP/Benefits to Clients:
 Reduce waiting time in salary distribution by 80%.
 Maintain proper financial records. Improved the compliance.
 Hassle free money distribution.
 100% reduction in accounting frauds.
 ZERO manual intervention.

Case Studies:
Biometric Based Offline ATM
The Client is world’s largest tea producer. They support local communities and provide all employees with a good environment to work in, they have around 90,000 employees in over 60 locations. They produce approximately 100 million kilograms of tea in a year. 90% of the worker based don’t have bank account so the worker salaries are paid through cash which requires manual work and prone to errors. The client is looking for a solution to automate the salary distribution process.

Business Drivers
 Increasing the productivity by automating the accounting process.
 Highly secured and easy way to withdraw money from ATM without the use of any card or pin but using a two factor authentication.
 Our solution uses dual authentication (Face and finger recognition) to verify / validate and authenticate the employee.

Scope
 Biometric validation and disbursement of cash through ATMs instead of the manual disbursement process.
 Wallet maintainability for every individual.
 Easy to track the financial details.
 24X7 Automated cash disbursement.
 Highly secured offline transaction processing using biometrics.
 Dispense the cash based on the balance in the account.
 Weekly updates with the central server.

Approach and Deliverables
 Implemented two factor authentication. Authenticate employees using face and finger.
 System can identify the person with worn out fingers or only if the partial faces are captured by the ATM camera.

Out Comes
 Reduce waiting salary distribution time by 80%.
 Maintain proper financial records. Improved the compliance.
 Hassle free money distribution.
 100% reduction in accounting frauds.
 ZERO manual intervention.
 

2. Forensica GPS



Desktop based verification and validation system for law enforcement officials. Forensic investigators and security personnel have a pressing need to resolve suspects' identities from any kind of photographic evidence that captures an image of a suspect’s face including video surveillance cameras. Forensica GPS, the new biometric solution from Insights of data, allows officers to identify individuals from low-resolution photos or videos. Forensica GPS features advanced facial recognition technology designed to increase the speed and enhance the accuracy of criminal investigations. Powerful Comparative and Analysis Tools Enhance Speed and Accuracy of Criminal investigations.

Features:
 Construct the full face if face is partially visible.
 Eliminate light illumination problem.
 Eliminate Pose variation problems.
 Map and measure the scars or marks are on the face.
 Easy to identify the person even with low resolution image or blurry image.

Benefits to the Clients:
 Identified the group of top 10 suspects after processing 2Mn facial records under near real time.
 Automated the identification of top 10 suspects on a daily basis by means of creating 3D face mesh and run the face recognition against the suspicious individuals.
 Increased investigation success rate by 40%.

Case Studies:
Suspects Identification – Law Enforcement

The client is a US based Law-Enforcement Agency. They were unable to complete the investigation of various cases as they have suspects facial images with poor resolution and partial faces. They are looking for a biometric solution which helps them in the investigation by identifying the suspects from their image database. The goal is to identify the top 10 suspects that has the closest match to the probe image.

Business Drivers
 Identification of suspects is very difficult to the agency with the poor resolution and partial face images.
 Investigators can generates three-dimensional avatars that can be rotated and viewed from any angle from partial 2D images.
 Help the investigators to resolve the identities of suspects face.

Scope
 Enabling 3D visualization of the given suspects partial face.
 Generates the frontal face view .
 Identify and form the cluster of top matched suspects.
 Give investigators the ability to correct suspect poses and fill in facial details difficult to perceive in the original images.
 Fast and accurate face identification resolution for criminal investigations, law enforcement professionals.
 Create visualization reports.

Approach and Deliverables
 Implemented face recognition Engine
a) To restructuring the partial 2D face into frontal face view.
b) To Run the face recognition against the database images.
 Created biometric face templates for each facial image that is stored in the Mongo DB document database.
 Visualization was performed using High charts.

Out Comes
 Identified the group of top 10 suspects after processing 2Mn facial records under near real time.
 Automated the identification of top 10 suspects on a daily basis by means of creating 3D face mesh and run the face recognition against the suspicious individuals.
 Increased investigation success rate by 40%.
 

3. Video-ID



A live video based recognition, where we maintain a database to store people face images with their metadata such as name, age etc.

Recognize and verify persons who came in front of the IP cameras installed at suited location. Examine the captured face against the database and perform an action based on the requirement.

Features:
 Advanced 3D facial recognition for smartphones and web connected digital devices.
 Real-time in field ID processing.
 Real-time watch list search.
 Advanced face biometrics for uncontrolled and low resolution facial Imagery.
 Forensic analysis and comparative tools to analyze and match single/multiple facial images in 2D/3D modes.
 Take necessary action when person identified positively or negatively
 Identify the person moving pattern.
 Can be easily identified even with partially visible face or face covered with something like cloth etc.

Customers:
Largest University in INDIA.
 

4. Mobile ID


An image based client-server model and embedded model with cross platform (Android and iPhone) where we maintain a database where are people faces images where enrolled with their metadata such as name, age etc. As per requirements.

When we probe picture to the system using mobile system check and validate against the database and responds back with result within few seconds.

Features:
 or – 45 degrees of pose variation is acceptable.
 On the fly recognition results.
 Advanced face biometrics for uncontrolled and low resolution facial Imagery.
 Take necessary action when person identified positively or negatively.
 Identify the person moving pattern.
 Can be easily identified even with partially visible face or face covered with something like cloth etc.

Customers:
Largest University in INDIA.