Liveness Detection from HMS Core ML Kit: Safer and Simpler

Search This thread

Vivi Jiang

Member
Dec 25, 2020
39
1
Facial recognition is used everywhere, such as for verifying your identity at the bank, clocking-in and –out for work, and even when entering some restricted buildings. On mobile phones, this technology allows us to unlock our phones and pay for things. And once integrated into apps, this technology facilitates easy sign-in and password change.

Behind the usefulness of facial recognition, however, lurks the risk that someone may use a fake face to trick and bypass this technology. The core concern for users of facial recognition is whether it is capable of telling whether a face is real or not.

The liveness detection service from HMS Core ML Kit overcomes this issue, and this explains why the APIs of this service have reached a great number of average daily calls and why it is well received among developers.

Following an upgrade to the liveness detection service, it will provide interactive biometric verification, aside from static biometric verification, helping improve user security and trust in facial recognition technology.


Identifying Each Fake Face Using Liveness Detection​

Facial recognition is a technology that enables a machine to recognize a person's face. Most facial recognition systems, however, can simply recognize a face in an image, but cannot accurately determine whether the face is of a real person. This has sparked the need for technology that can automatically distinguish fakes faces from real ones, to prevent spoofing attacks.

Such technology can be realized using the liveness detection algorithm. It can detect such fake faces as those printed out, displayed on an electronic device, or disguised as a silicone mask or 3D portrait, to prevent fake face attacks.

This technology is widely used in finance, public affairs, and entertainment, which also makes its application challenging. For example, the expectations for liveness detection vary depending on the device, people, and environment involved, meaning that this technology needs to be constantly upgraded.

Improving User Experience with Interactive Biometric Verification​

ML Kit will offer the interactive biometric verification capability to strengthen the flexibility of its liveness detection service. An app with this capability can prompt a user to do either three of the following actions: blink, open their mouth, turn their head left, turn their head right, and stare at the camera. If the required action is not detected, the face will be deemed fake.

With the deep learning model and image processing technology, liveness detection is useful in many scenarios by providing prompts that indicate the lighting is too dark or bright, a mask or sunglasses are blocking the view, and the face is too near to or far from the camera. This ensures that the whole liveness detection process is efficient, secure, and user-friendly.

The liveness detection capability can help perform remote identity authentication in fields such as banking, finance, insurance, social security, automobile, housing, and news. It is a cost-effective solution thanks to its simple steps for performing remote identity authentication and service access.

Following an upgrade, liveness detection will offer two methods of authentication: static biometric verification and interactive biometric verification.

  • Static biometric verification has received some groundbreaking updates, by utilizing data in more than 200 scenarios. Such data is collected through cooperation with the data company, which makes this method useful in almost every scenario where it is needed.
  • Interactive biometric verification will come with a well-developed SDK, framework for calling its algorithms, and reference UI, which simplify integration.
These two methods can be used flexibly in situations such as authenticating user identity during insurance purchase, in the anti-addiction system for a game, during the real-name registration for SIM cards, and during the activation of a live-streaming function or reward permission.

By leveraging AI, ML Kit will make liveness detection more secure, accurate, and versatile to deliver a safer and more user-friendly experience for business and individual users.

To know more about liveness detection, please refer to its official document.
 

Basavaraj.navi

Senior Member
Dec 4, 2020
86
3
29
Bangalore
Facial recognition is used everywhere, such as for verifying your identity at the bank, clocking-in and –out for work, and even when entering some restricted buildings. On mobile phones, this technology allows us to unlock our phones and pay for things. And once integrated into apps, this technology facilitates easy sign-in and password change.

Behind the usefulness of facial recognition, however, lurks the risk that someone may use a fake face to trick and bypass this technology. The core concern for users of facial recognition is whether it is capable of telling whether a face is real or not.

The liveness detection service from HMS Core ML Kit overcomes this issue, and this explains why the APIs of this service have reached a great number of average daily calls and why it is well received among developers.

Following an upgrade to the liveness detection service, it will provide interactive biometric verification, aside from static biometric verification, helping improve user security and trust in facial recognition technology.


Identifying Each Fake Face Using Liveness Detection​

Facial recognition is a technology that enables a machine to recognize a person's face. Most facial recognition systems, however, can simply recognize a face in an image, but cannot accurately determine whether the face is of a real person. This has sparked the need for technology that can automatically distinguish fakes faces from real ones, to prevent spoofing attacks.

Such technology can be realized using the liveness detection algorithm. It can detect such fake faces as those printed out, displayed on an electronic device, or disguised as a silicone mask or 3D portrait, to prevent fake face attacks.

This technology is widely used in finance, public affairs, and entertainment, which also makes its application challenging. For example, the expectations for liveness detection vary depending on the device, people, and environment involved, meaning that this technology needs to be constantly upgraded.

Improving User Experience with Interactive Biometric Verification​

ML Kit will offer the interactive biometric verification capability to strengthen the flexibility of its liveness detection service. An app with this capability can prompt a user to do either three of the following actions: blink, open their mouth, turn their head left, turn their head right, and stare at the camera. If the required action is not detected, the face will be deemed fake.

With the deep learning model and image processing technology, liveness detection is useful in many scenarios by providing prompts that indicate the lighting is too dark or bright, a mask or sunglasses are blocking the view, and the face is too near to or far from the camera. This ensures that the whole liveness detection process is efficient, secure, and user-friendly.

The liveness detection capability can help perform remote identity authentication in fields such as banking, finance, insurance, social security, automobile, housing, and news. It is a cost-effective solution thanks to its simple steps for performing remote identity authentication and service access.

Following an upgrade, liveness detection will offer two methods of authentication: static biometric verification and interactive biometric verification.

  • Static biometric verification has received some groundbreaking updates, by utilizing data in more than 200 scenarios. Such data is collected through cooperation with the data company, which makes this method useful in almost every scenario where it is needed.
  • Interactive biometric verification will come with a well-developed SDK, framework for calling its algorithms, and reference UI, which simplify integration.
These two methods can be used flexibly in situations such as authenticating user identity during insurance purchase, in the anti-addiction system for a game, during the real-name registration for SIM cards, and during the activation of a live-streaming function or reward permission.

By leveraging AI, ML Kit will make liveness detection more secure, accurate, and versatile to deliver a safer and more user-friendly experience for business and individual users.

To know more about liveness detection, please refer to its official document.
Is it free or paid?