Find Us

123 Main Street
New York, NY 10001

Monday—Friday: 9:00AM–5:00PM
Saturday & Sunday: 11:00AM–3:00PM


Behavioral biometrics: gait analysis – voice ID

In this series, we will be presenting the various types of biometrics and explore their use for authentication and beyond. In our latest blog post, we spoke about mouse use characteristics and analyzed their use cases and advantages for both companies and end-users alike. Now, let’s look into voice ID and gait analysis.

Gait analysis

Gait analysis has been interestingly used in the sportswear industry for decades. It’s not only a fancy shoe store addition but can represent a biometrics authentication factor. The way we move or walk has also been used for various medical reasons in physical therapy and rehabilitation. The aim is to help patients suffering from different diseases or inabilities. In the medical circles, constant advancements are achieved, and recently, there are also references on normal or pathologic gait.

How does it work?

Kinematics, movement patterns, kinetics and the forces involved in producing those movements are analyzed, and data is generated. On mobile devices, the gyroscope is used to gather information and fitness trackers can detect various movements and create individual profiles.

The technology analyzing specific traits of how people walk or move is already being used for other purposes than a medical diagnosis. Smart vehicles are one example. The technology is used to identify the person when entering a vehicle. If the gait analysis returns a known behavior, the car will automatically open, enabling a futuristic, keyless experience.

With time, specifically for authentication requiring a lower threshold of security, gait analysis apps will be a very convenient and nonintrusive method to verify identity. 

However, there are some disadvantages that will not make this technology gain momentum in the near term. The hardware needed to authenticate a user makes it a no go. Also, the accuracy is very low compared to other biometrics widely used, such as face ID or typing biometrics.

Voice ID 

Each individual can be identified based on the unique biological characteristics of the voice. speech recognition, known as ASR(automatic speech recognition) was first introduced in 1952 and allowed a machine to recognize single digits spoken by a person.

Today, software programs allow people to train ASR and accurately translate speech to text. This biometrics technology is not only circumventing the user experience problem of hard-to-remember-passwords but has also enabled the visually impaired to make use of computers. 
The use cases vary but are not limited to usage in medical devices and in the military. Also, the US has used this technology for more than ten years to ensure that prisoners incarcerated in their homes were actually there.

In the past years, due to improved accuracy of ASR, voice recognition has also been frequently used to verify the identity of users. A successful use case is an authentication for clients who contact call centers. Several banking institutions have also implemented this technology to verify the identity of their customers.

Passwords alone are not safe anymore. There is a multitude of use cases where they can’t be implemented or the level of security requires an additional security factor. Biometrics represent a viable alternative and in the next few years, most experts predict that they will be used more frequently.

Voice ID recognition has excellent potential to become one of the next generations’ preferred authentication methods, competing with other nonintrusive and very accurate factors such as typing biometrics