Liveness detection

Introduction

The National Crime Record Bureau (NCRB), in its report, recorded 50,035 cybercrime cases in 2020, and it is alarming to see that the numbers have surged exponentially ever since the pandemic. This opened the gates to remote onboarding and the use of facial recognition algorithms to distinguish users’ unique features, enhancing their onboarding experience.

On the flip side, this has also given scamsters and fraudsters ample opportunities to manipulate and employ technologies for wrongful gains. It’s now an absolute must for companies to invest in updated security measures; even more so when it comes to user verification and customer onboarding.

Liveness detection

Liveness detection is a technique used to identify any spoof attempts in determining whether a biometric sample is a live human being or a fake representation. This is achieved through AI-based algorithms, that interpret data collected from sensors to assign a liveness score. The liveness score is then passed on to the organisation as the status feedback to make final decisions.

Getting in pace with the post-Covid world, banks and other businesses have now moved to mobile onboarding. Today, liveness detection is particularly a useful biometric module for mobile onboarding and authentication for the following reasons-

  • Ease of Integration & Compatibility
  • Intuitive & Convenient
  • Accuracy

Mobile onboarding allows a prospective customer to open an account without visiting a branch. Liveness detection makes this identity proofing process more robust and secure so that, new customers can then use their biometrics in place of passwords for more secure and convenient authentication. While facial recognition is an ideal biometric modality for mobile applications,

 Types of liveness detection 

Active liveness detection. This involves the customer in performing a task – users might be prompted to blink, smile, or move their device during a facial recognition capture. Users are also hence completely aware of the liveness detection measures applied.

Passive liveness detection. This technique uses deep learning and doesn’t need customer interaction. Passive liveness detection can be activated in the background and relies on algorithms that can identify and record strategic data points and micro muscle movements based on the image contents. It can detect photo of a mobile photo, photo of a video, deepfakes or even latex masks.

How it Works

Step 1: Client application through which liveness detection API is called

Step 2: Single image frame is then captured and uploaded

Step 3: Invoking APIs to run the liveness detection algorithms: Enabling real-time liveness analysis with very little API latency

Step 4: Liveness Status Feedback: Allowing businesses to control the threshold of success or failure

(*The image captured by IDcentral Liveness Detection can also be used for other facial recognition services such as Face Search and Face Trace)

 Advantages of passive liveness detection:

  • Fast: Takes less than 3 seconds: Passive liveness detection technology can run in the background without the user realizing its occurrence -“security through obscurity”
  • Frictionless: This approach requires absolutely no participation of the user proving better user experience and customer conversions (<1% drop-off rates).
  • Cost Efficient: Reduces infrastructural and onboarding costs, unlike other solutions that require high bandwidth connections and large-scale server deployments for processing.
  • Works for Everyone: IDcentral’s AI-Based algorithms deliver unbiased results across demographics including race, gender, and age.
  • Flexible: Works with all modern smart devices and web cameras (Stores no data and offers scalability).

Comparison

Liveness Detection  Passive Liveness Active Liveness 
USER EXPERIENCE Requires no action by the user, resulting in lower abandonment, less time & effort Requires users to perform a “challenge-response” that adds time and effort to the process
SOFTWARE REQUIREMENTS API/SDK Integration API/SDK Integration
IMAGE ANALYSIS Real-time AI-Based processing of images captured at an instant Requires analysis of a series of images of video to detect
the requested movement.
BANDWIDTH REQUIREMENTS Low High
SPEED Real-Time Active liveness increases user effort, resulting in a longer liveness check.
COST
ROBUSTNESS TO SPOOFING High immunity to spoofing Very high immunity to Spoofing
INTEGRATION EFFORT Supports both API & SDK Integration Requires front-end to capture video separately, and stream processing to then transmit the video
USER DROP-OFF RATE  <1%  5% to 10%
CONNECTIVITY Works with & without an active internet connection Works with & without an active internet connection

 USE CASES

Remote Customer Onboarding– Enabling new customers to sign up digitally thereby increasing the opportunity to acquire more users.

Biometric-based device login– With liveness detection, users can get instant access without carrying tokens or cards.

Attendance- Enabling a quick, effortless and secure attendance system

Digital Banking- Build trust amongst your customers by authenticating them for high-value transactions and other important service requests.

Online Check-in and Check-out- Enhance your customer experience and queueing strategy at Airports, hotels, lounges etc.

How Liveness Detection Fits into Your Onboarding Process

Step 1:

Document Check: Is the ID document authentic and valid?

Step 2:

Comparison Check: Is the person holding the ID the same person shown in the ID photo?

Step 3:

Liveness Check: IDcentral performs liveness detection using advanced selfie algorithms to ensure that the user is real & is physically present at the time of the detection.

Step 4:

Definitive Answer: IDcentral delivers a definitive yes or no answer in seconds.

Future to Liveness Detection 

The world is changing rapidly, and user experience has emerged as the vital essence for the scaling of any business. Matching the customer’s expectation with the onboarding is challenging, especially when many organizations are already using advanced technology to offer an excellent onboarding experience to their customers.

IDcentral offers both active and passive Liveness detection solutions, providing real-time & accurate results tailored for the global clientele. Offering the highest accuracy rates in the industry (97%), it detects spoofing attacks via mobile video & photo, print of photos masks or even latex masks. This AI-powered Liveness detection tool offers a unique solution for anchoring a person’s digital identity. It is easy to integrate and guarantees a faster, seamless, and fraud-free onboarding resulting in better first impressions and long-term business value. The AI algorithms smartly determine whether it interfaces with a live human being or an inanimate spoof which significantly mitigates the vulnerability that comes with biometric systems against spoof attacks.  

Real-time biometric verification to protect you from fake customers

Request demo

IDC logo

Differentiate the good from the bad. Filter out fraudulent customers before they impact your revenues.