Inspired R&D

What's in a name? For us, it's R&D.

Biometrics research and development is at the heart of ID R&D. It’s where we invest the most resources, failing fast and succeeding sooner to stay a step ahead.

Our team is dedicated to advancing the science of biometrics and using the full power of AI to transform the way people securely access information and interact with a growing number of connected devices.

Our team is made up of industry veterans, PhDs, award-winning developers and engineers who have deep expertise in speech, biometrics, AI and machine learning. The team regularly presents biometrics research and participates in challenges and competitions, achieving top rankings in NIST ASR, NIST Language ID, ASVSpoof and other leading benchmark evaluations.

Best-In-Class Voice Biometric Liveness

ID R&D is a clear leader in the largest voice biometric anti-spoofing challenge to date

ASVspoof 2019 challenge results

Our biometrics research focuses on

  • Voice Biometrics using Artificial Intelligence, including Deep Neural Networks and Convolutional Neural Networks
  • New inventions such as our patented p-vector algorithms and optimized x-vector algorithms for feature extraction
  • Anti-spoofing to protect voice and face biometrics from fraud attacks
  • Acoustic Event Detection using Deep Neural Networks and signal processing analysis of sounds to detect and alert on non-speech events of interest

From Research to Reality - Industry Firsts

Speed and Accuracy
ID R&D is first to market with x-Vector technology, resulting in the fastest and most accurate text-independent verification solution on the market.
Passive Facial Liveness
ID R&D’s IDLive Face is the world’s first truly passive facial liveness product, requiring no actions by the user.
Reduced Footprint for Mobile
Using x-vector, we have decreased the size of the voice biometric engine and industry-leading voice anti-spoofing to under 20M, much less than lower performing i-Vector solutions.
Language Support
ID R&D uses the latest artificial intelligence algorithms in our biometrics SDKs and is trained to be independent of language. Other solutions require data from specific languages.
Omni-channel Authentication
Enable voice verification across multiple channels and devices without re-enrollment.

Data Collection and Performance

Quality data is critical to machine learning at scale. The ID R&D team has collected a vast amount of data from universities, the scientific community, crowdsourcing initiatives and internal projects. Additionally we acquire publicly available commercial data sets from organizations such as NIST and IEEE. Our ability to source large amounts of properly collected and labeled data has been an important factor in training and improving our biometrics research, algorithms and biometrics SDKs.

Facilitating Innovation

ID R&D’s products are delivered as flexible biometrics SDKs. One or more products can be integrated to develop a custom solution or enhance an application with additional functionality.

Our solutions bundle ID R&D products with third party products and professional services for integration and customization. Our solutions are intended to provide a blueprint for the application of one or more products to solve a specific industry need or business problem.

Biometrics SDKs are available for iOS, Android, Linux, or Windows, and as a Docker image. If you have a custom device or need to integrate at a chip-level, contact us. We can deliver customized packages with small footprints.

If you have a problem to solve or an idea you think we can help bring to reality, let's talk.