When it comes to leverage a human trait for biometric applications, options that generally come to mind are fingerprints, iris, face, voice, etc. Recognition methods based on these human characteristics have been largely adopted and implemented throughout the world and these are the biometric methods we generally come across in everyday life. In offices, fingerprint attendance systems, fingerprint door access and face/fingerprint login in laptops are not something that surprise us anymore. Biometrics has already replaced traditional identification and authentication practices in many applications. Governments of various countries have accepted biometrics as mainstream recognition method and completely replaced traditional card and document based methods of identity verification.

Traits that are leveraged for biometric applications are either picked form physiological characteristics (which are present in or on human body, like friction ridges, iris patterns, retina, DNA, etc.) OR from behavioral characteristics (habits that develop over time and are considered permanent, like signatures, gait, typing rhythm, etc.). A biometric characteristic may also fall in both the categories, for example voice. Voice is resulted out of anatomical structure of vocal chords as well as habitual way of speaking, making voice an individual characteristic. While these conventional methods are used the most for biometric applications, there are other human traits as well that do not get enough attention, but have a lot of scope for research and specialized deployment. Let’s have a look at these 5 unconventional biometric traits, which are not considered mainstream, but have potential to become one.

While we all have ability to recognize others by perceiving their gait, physical structure, and face, it is not common among human beings to recognize each other by their body Odor, however, animals like dogs have extraordinary capability to recognize other animals as well as humans by their scent. This biological ability gave a hint that body odor can be an individual characteristic despite human inability to differentiate it. Where human ability falls short, machines are leveraged to fill the gap and this is where biometrics came to the picture. Researchers at the Polytechnical University of Madrid have been working on a biometric system that can recognize people by their unique body odor. Body odor of each individual differs slightly and this minute difference can be used to positively identify that individual. The research was carried out by the Group of Biometrics, Biosignals and Security (GB2S) at the Universidad Politécnica de Madrid (UPM) in collaboration with Ilía Sistemas SL.

While we all have ability to recognize others by perceiving their gait, physical structure, and face, it is not common among human beings to recognize each other by their body odor, however, animals like dogs have extraordinary capability to recognize other animals as well as humans by their scent. This biological ability gave a hint that body odor can be an individual characteristic despite human inability to differentiate it. Where human ability falls short, machines are leveraged to fill the gap and this is where biometrics came to the picture. Researchers at the Polytechnical University of Madrid have been working on a biometric system that can recognize people by their unique body odor. Body odor of each individual differs slightly and this minute difference can be used to positively identify that individual. The research was carried out by the Group of Biometrics, Biosignals and Security (GB2S) at the Universidad Politécnica de Madrid (UPM) in collaboration with Ilía Sistemas SL.

An individual’s body odor comprises of chemical compositions that can be analysed and results can be used to identify that individual. Primary odor of an individual contains components that stay stable over time and does not change with factors like food or environment. Secondary odor contains components that are present due to diet and environmental factors, while tertiary odor contains factors which are present due to external factors like soaps and perfumes. Since primary body odor of an individual stays stable over time, it can be used to identify that individual. For the purpose, sensors are used for collecting the human scent sample data. Sample data is taken through odor biometric systems and a unique template is generated. This template is stored in a biometric database and can be used to verify identity of the individual. Odor biometrics makes sure that scents from external sources like perfumes and deodorants cannot mask actual body odor.

The bright side of odor recognition is that it is a completely non-invasive and can be performed without user consent. So subjects can be screened unaware that their body odor is being leveraged to verify their identity.

Odor biometrics can make a difference by laying security and safeguarding national interests from criminals and terrorists. Increasing crime and terrorist incidents have made identification of these anti-social and anti-national elements a necessity. Identity of criminals and terrorists can be established with an odor recognition system, which can later be used to surveil them, without their knowledge.

Natural calamities like earth quake and collapse of buildings leave people buried in debris and it becomes impossible to locate them. Artificial or electronic nose can be used to figure out location of people buried in debris and rescue them.

Odor biometrics can be implemented on airports and border checkpoints to verify identify of the people crossing the national borders.

Artificial or electronic nose can digitize body odor and transfer it to a remote location over the internet. This can be useful in applications like tele-surgery, where surgeons may need to identify certain smells.

It will be interesting to see if this method makes it to mainstream identification and authentication applications and how does it fare.

We all have flaunted our reminiscence ability by recognizing a movie star just by his or her distinctive style of walking, even though his face is not visible. We can also recognize people we know by the way they walk, without looking at their face. This ability comes naturally to human brains; it can recognize a moving pattern, remember it and effortlessly recall it. When this ability is given to computers, it is called gait recognition system and area of study is called gait biometrics. A person’s gait is dependent on several biological, environmental and habitual factors. Being dependent on several factors, gait becomes a unique behavioral characteristic, which can be used for personal identification. Gait recognition system is basically a system that has ability to map patterns of human motion. Gait analysis is also used to study animal locomotion. The mapped data can be stored and gait biometric system can recognize the same pattern of locomotion by capturing and processing a new gait signature.

A gait recognition system can make use of video feed as well as sensor data to process and map pattern of an individual in motion. Raw data requirements depends on the input requirements of a recognition system, for example, some recognition systems may just need video feed, others may require sensor data as well. Gait biometrics can be useful in personal identification in scenarios where other biometrics fail, for example: in a moving crowd. User consent is not required and often not taken when it is used for surveillances purposes. Gait analysis is among the least invasive, most passive and most undetectable techniques. Gait signature analysis does not require cooperation from the subject as it can be performed from a distance without subject’s knowledge. Enrolling and putting people on surveillance without their consent is one the potential privacy concern often expressed regarding gait recognition.

As the area of biometrics advances, newer trends are emerging in gait analysis. 3D gait analysis is one of these trends, which promises of very accurate measurements of human movement. Many athletes, runner and people suffering from injuries go for a 3D gait analysis to understand how they can improve performance and recover from injury.

Human ear offers a fair possibility to become a biometric identifier for personal identification. Its unique shape contains curves and lobe that differ significantly from person to person. Structure of human ear stays stable with age, unlike face, which can change not only with age, but also with different expressions. Stability in ears’ shape becomes more advantageous than facial recognition for personal identification. This uniqueness of shape of human ear can be mathematically mapped and used to identify a person.

Ear biometric methods are based on processing of digital images of ear. A number of methods have been worked upon to calculate different features of ear mathematically and work is still underway to improve ear recognition. Ear prints have been successfully used in forensics to solve crimes in the past. A perpetrator leaves ear prints on doors and windows when trying to listen to any activity inside the room. Successful usage of ear identification at forensics gave it enough reputation to use it in biometrics.

Ear shapes were already good enough to be unique, but NEC Corporation, which is a Japanese multinational technology company, went the extra mile with ear identification. NEC announced a new biometric personal identification technology that uses the resonation of sound determined by the shape of human ear cavities to distinguish individuals. In a layman’s language, this technology basically measures the acoustic signals reflected by human ear cavities, so it is not dependent ear print or image analysis. NEC claims that this recognition system of distinguishing individuals based on acoustic characteristics has greater than 99% accuracy.

We all use gestures to emphasize, express or communicate familiar or prearranged notions or signals. Gestures make things easy when communicating with people, but what if same gestures could be used to communicate with computers and equipment? Well, it is now actually possible to use gestures to perform man – machine interaction. It has become possible with implementation of gesture recognition by leveraging techniques to understand movement of hands and perform predefined commands with it. This mathematical interpretation of human motions by a computing device is called gesture recognition. Gesture recognition, along with facial recognition, voice recognition, eye tracking and lip movement recognition are components of what developers refer to as a perceptual user interface (PUI). The goal of PUI is to enhance the efficiency and ease of use for the underlying logical design of a stored program, a design discipline known as usability.

Gestures are already being leveraged by computing and smartphone devices to accept user inputs. For example, many smartphones and tablets have ‘gesture typing’ feature, in which users can either write the letter with their finger or slide it over the individual letters to write a word. In a virtual reality environment and games, systems can understand user gestures and respond accordingly. It makes virtual reality systems and games more interactive and immersive. A system can make use of sensor equipped wearable or an outfitted device with a camera or radar signal generator to collect motion data and interpret user gestures, for example, Google’s gesture recognition project called Soli makes use of radar signals to collect and interpret motion data. Microsoft’s Kinect is another example, in which user motions is interpreted with the help of an external device equipped with RGB and IR depth finding camera.

User gestures can also be used to authenticate a user. Microsoft’s latest operating system Windows 10 comes with ability to login by drawing gestures on a picture. This feature is called ‘Picture Password’, which is a marketing term by the company. User can simply select a picture and draw a gesture on it and save it to login with the same gesture later on.

Pattern made by lip grooves offers a certain degree of uniqueness and can be used to identify a person. Study of lip patterns for personal identification is called Cheiloscopy. Pattern made by lip grooves can significantly differ from person to person as it offers a huge possibility of variations. Lip prints have been used in forensics for a long time and have had evidentiary value for positive identification of a person. Lip patterns are recognized by the location, pattern and type of grooves they have. Lips pose behavioral characteristics as well as they contribute substantially to facial expressions and smile.

Despite being unique and an individual feature, technological advancement in the area of lip prints is not as great as iris or fingerprint recognition. There is not automated system to digitize lip pattern data and run a match. Cheiloscopy has mostly served forensics as a personal identification aid. This area still left with enormous scope of development.

What is conventional is easy. Easy to adopt, implement and get support for. Going for unconventional requires courage and commitment. Same applies in case of unconventional biometrics recognition methods. Choosing unconventional biometrics for user identification and authentication application will require a lot of commitment. System even might not work up to your expectations. Some unconventional biometrics are gradually getting momentum and being adopted in different applications, for example: gait biometrics are being used by security and law enforcement agencies; on the other hand, gesture biometrics is in use with mobile computing devices for user authentication, user input, etc. However, some other unconventional biometric recognition methods are still in development phase and not ready for mass deployment, e.g. body odor identification. Despite their infancy, unconventional biometric methods have a lot of potential and can be useful where other biometric methods reach their limit.