Since the iPhone 5s was equipped with fingerprint recognition, fingerprint recognition has become the standard for smartphones in a few years. In addition to fingerprint recognition, face recognition, iris recognition, and speech recognition are all biometric technologies that increase speed. What are the characteristics of these technologies? Why are Apple, Google, Huawei, Samsung, etc. optimistic about face recognition?
Biometric technology is primarily a technique for identity authentication through biometrics such as measurable body or behavior; and biometrics are the only physiological features or behaviors that can be measured or automatically identified and verified.
Biometric market sizeAccording to the China Foreground Industry Research Institute, the global market for biometrics grew at an average annual rate of 21.7% from 2007 to 2013. In 2015, the global market for biometrics technology will reach US$13 billion, and it will reach US$25 billion in 2020. The average annual growth rate will be about 14% in five years.
From 2015 to 2020, the market size of each segment is: fingerprint (73.3%), voice (100%), face (166.6%), iris (100%), and others (140%).
Four biometric technologies
Fingerprint recognition
Fingerprint recognition is currently the most widely used biometric technology, mature technology and low cost, widely used in attendance, access control and other identification. However, it is easy to copy, and the fingerprint is worn to affect the recognition accuracy.
Fingerprint recognition refers to the identification by comparing the detailed feature points of different fingerprints. Because each person's fingerprint is different, there is a clear difference between the fingers of the same person, so the fingerprint can be used for identification.
The advantage of my fingerprint identification technology is that it is more convenient to apply, long-term application, and high recognition. But like many plots in large movies, fingerprints are a feature that can be easily stolen and copied, and the security is low. In addition, the stability of fingerprint features is poor, peeling, epidermal tweezers, dry and wet conditions will affect the application of fingerprints.
Speech Recognition
Speech recognition is authenticated by analyzing the unique characteristics of speech. Its device range is large and easy to install, but the recognition accuracy is low, it may be deceived by recording, and it is susceptible to background noise, physical condition, emotion and other factors.
Speech recognition technology will also mature before 2030, but Yole market and technology analyst Guillaume Girardin emphasizes that it now requires "very good microphones, such as Vesper Technologies, which perform the necessary noise cancellation and complex advanced beamforming. Even so, for financial transactions, this technology may still be less secure because many professionals may emulate other people's voice patterns.
Face recognition
Compared with other biometric methods, face recognition has the advantages of naturalness and undetectedness. Naturalness means that the recognition method is the same as that used by humans for individual identification. Humans also distinguish and confirm other individuals by observing information such as faces and sounds. Therefore, fingerprint recognition, iris recognition, and the like are not natural. The undetected feature makes the identification method difficult to make people conflict. The visible image can be used to obtain facial image information, while the fingerprint recognition or iris recognition needs to use electronic pressure sensor or infrared to collect fingerprint and iris image. In the process of collection, the identity information is It is possible to be copied.
Iris recognition
Iris technology is a recognition technique that utilizes the texture features of the iris of the human eye. The iris is a fabric-like ring in the pupil of the human eye. The iris recognition uses the features of the iris region (rings, wrinkles, spots, crowns) in the human eye image to form a feature template, and the identification is completed by comparing the feature parameters. The method has high recognition accuracy and is not easy to counterfeit, but the related equipment is expensive.
Iris recognition is a more accurate one in terms of technical indicators, but it needs to be imaged by infrared or visible light to obtain images. The user has low degree of cooperation and high psychological rejection, so it is not suitable for ID card. Large crowd application. In addition, products such as Miele that appear on the market can change the characteristics of the iris and are easier to copy.
The giants have laid out face recognition technologyAccording to Girardin, Microsoft and Apple are working on face recognition technology. Microsoft confessed that it will be importing face recognition, and Girardin believes that Apple's previous acquisition of FaceShift, a company focused on tracking faces on the move (video-based facial recognition technology), is the best evidence to cross the field. . Girardin pointed out that although Samsung has taken the lead in iris recognition, he expects that the development of face recognition technology will be faster and better than iris recognition.
1. Apple first began to face the field of face recognition in 2010, especially the recent acquisition of artificial intelligence and machine learning face scanning companies. In September 2010, the acquisition of Swedish algorithm company Polar Rose, the acquisition of Prime Sense in 2013, the acquisition of machine learning and image recognition company PercepTIo in September and November 2015, and the motion capture company Faceshift, Apple in January 2016 In August, they acquired EmoTIent and Turi respectively. EmoTIent is a face recognition system company and Turi is a machine learning company. In 2015, the company obtained a patent for face recognition unlocking device, specifically protecting the use of infrared sensors, in line with the needs of face recognition.
Apple has obtained a patent for face recognition unlocking equipment in 15 years. Based on the recent development of the industry chain, we expect to put it into use with great probability.
2. Samsung applied for patents for face recognition devices, algorithms and machine-readable media on December 29, 2010; applied for equipment and algorithm patents for image recognition feature extraction on June 19, 2004;
3. Huawei applies for at least 10 face recognition related patents, such as methods, devices, and terminals for determining the identity of a face in a face image;
4. In July 2011 and October 2012, Google Inc. acquired face recognition system company PittPatt and Ukrainian face recognition company Viewdle. As of June 2016, it has applied for 21 face recognition related patents.
5. Facebook acquired Tele Aviv, an Israeli face recognition platform company, in June 2012 and applied for two patents related to the online suggesTIons for images on online social networks. In 2014, DeepFace, a deep face recognition learning system, was developed.
6. Amazon and Microsoft also applied for 7 and 6 facial recognition patents respectively.
DJ Subwoofer,DJ Speaker Price,Portable DJ Speaker,12 Inch DJ Speaker Price
Newmax Electronics Co.,LTD , https://www.fspeaker.com