Although AI technology is the hottest development, it is biometric technology that is faster than it enters the mature stage. Dharma believes that biometrics will enter the stage of large-scale application in 2019.
In 2018, the Internet industry ushered in one after another “cold winter”, but technology still has the magic of re-ignition. On January 2nd, Alibaba Dharma Institute released “2019 Top Ten Technology Trends”, covering smart cities, digital identity, autonomous driving, graph neural network systems, AI chips, blockchain, 5G and other fields. It can be predicted that AI is still the hottest direction in the technology industry in 2019.
Dharma believes that biometrics will enter the stage of large-scale application in 2019. In the past few years, most people have become accustomed to go out without cash, and in the future, the era of going to the world without ID cards will come.
Biometrics is a technology for identity authentication through human biometrics. Human biometrics are usually unique, can be measured or automatically identified and verified, hereditary or lifelong, so biometric authentication technology is more traditional. Certification technology has great advantages. The biometric system samples the biometric features, extracts their unique features and converts them into digital codes, and further groups the codes into feature templates. As the cost of microprocessors and various electronic components continues to decrease, the accuracy is gradually increasing, and biometric systems are gradually being applied to commercial authorization control. Since the human body features the inherent non-reproducibility of the human body, this biometric key cannot be copied, stolen or forgotten, and biometric identification technology is used for identity identification, security, reliability and accuracy.
According to the US consulting firm Transparency Market Research, the global biometrics market will grow to $23.3 billion in 2020, with a compound annual growth rate of 15.7%. The biometrics market is experiencing rapid growth.
According to the “The Future of Biometrics” published by Acuity Market Intelligence, the US think tank, the fingerprint recognition share reached 58% in the global biometric market structure in 2015, and the face recognition share was 18%, followed by The emerging iris recognition has a share of 7%. In addition, there are palmprint recognition similar to fingerprint recognition, as well as voiceprint recognition and vein recognition.
Face recognition is a kind of biometric recognition technology based on human facial feature information. It uses a camera or camera to collect images or video streams containing faces, and automatically detects and tracks faces in the images, and then detects them. The face is a series of related techniques for the face, usually called portrait recognition and face recognition.
Nowadays, face recognition technology is very mature, and the domestic industrial chain is also becoming more and more perfect, such as face recognition technology, face recognition technology, face++ technology, business technology, and video recognition.Nowadays, the application of face recognition has not only been limited to business places, but it has gradually penetrated into daily life in various forms.
Each fingerprint has several unique measurable feature points, each with approximately seven features, and one of the ten fingers produces a minimum of 4,900 independently measurable features. The fingerprint recognition technology extracts the feature values from the fingerprint by analyzing the feature points that can be measured by the fingerprint, and then performs authentication.
At present, China’s second-generation ID card has achieved fingerprint collection, and all major smartphones have realized fingerprint unlocking functions. Compared with other biometric technologies, fingerprint recognition has been widely used in consumer electronics, security and other industries. Through time and practice testing, technology is constantly innovating.
Palm print recognition
Palm print refers to various ridge features on the palm surface between the wrist and the finger, such as main lines, wrinkles, fine texture, ridge tips, bifurcation points, and the like.The shape of the palm print is controlled by genetics, and even if the epidermis is peeled off for some reason, the new palm line remains in its original structure. Everyone’s palm lines are different. Even if they are twins, their palm prints are just similar, not exactly the same. And it can still be clearly identified in low resolution and low quality images.
The structure of the human eye consists of the sclera, iris, pupillary lens, retina and the like. The iris is an annular portion between the black pupil and the white sclera that contains many detailed features of interlaced spots, filaments, crowns, stripes, crypts, and the like. Moreover, after the iris is formed in the fetal development stage, it will remain unchanged throughout the life course. These characteristics determine the uniqueness of the iris characteristics and also determine the uniqueness of the identity. Therefore, the iris feature of the eye can be used as an identification object for each person.
Iris measurement technology can read 266 feature points, while other biometric techniques can only read 13-60 feature points. According to Fujitsu data, the iris recognition error recognition may be 1/1500000, while the Apple TouchID error recognition may be 1/50000, and the iris recognition accuracy is up to 30 times that of the current fingerprint scheme. Iris recognition is a non-contact identification, which is very convenient and efficient.
The generation of human language is a complex physiological and physical process between the human language center and the vocal organs. The vocal organs used by people in speech–tongue, teeth, throat, lungs, and nasal cavity vary greatly in size and shape. , so the soundprints of any two people are different. This also makes voiceprint recognition a form of authentication.
Compared with other biometric features, the advantages of voiceprint recognition are: (1) the soundprint extraction is convenient, can be completed unconsciously, so the user’s acceptance is also high; (2) the acquisition of voice recognition cost is low, use Simple, one microphone can be used, no need for additional recording equipment when using communication equipment; (3) Suitable for remote identity verification, only need a microphone or phone, mobile phone can be remotely logged through the network (communication network or internet) (4) The algorithm for voiceprint recognition and confirmation is low in complexity; (5) with some other measures, such as content recognition through speech recognition, the accuracy can be improved. These advantages make the application of voiceprint recognition more and more popular among system developers and users.
One way of the vein recognition system is to obtain a personal vein distribution map through a vein recognition instrument, extracting feature values from a vein distribution map according to a dedicated comparison algorithm, and another method to obtain an image of a finger, a palm, a dorsal vein of the hand through an infrared CCD camera, and a vein. The digital image is stored in a computer system to implement eigenvalue storage.When the vein is compared, the vein map is taken in real time, and the advanced image filtering, image binarization and refinement methods are used to extract the features of the digital image, and the complex matching algorithm is used to match the vein characteristic value stored in the host to match the individual. Identification, identification.
Specifically, in terms of AI technology, if the AI enters the reality from the laboratory in 2018, then 2019 will open a new starting point for comprehensive cooperation between humans and AI. Alibaba Dharma believes that real-time speech generation on mobile devices may not be distinguishable from live voice, and even voice AI will pass the Turing test in some specific conversations. In the city, there will be more and more talking facilities. Among them, biometric technology will enter the large-scale application phase in 2019. With the rapid spread of 3D sensors and the integration of multiple biometrics, each device can “see” and “listen” smarter. Biometrics and in vivo technologies will reshape identity and certification, and in the future, digital identity will become the second identity card for people.