Cutting edge technology

AI vision era Biometric technology empowers smart finance

Introduction: Recently, in the large-scale scientific and technological variety show "Intellectual Future", Alipay developed the face artifact for the first time, successfully found the target mission among 500 masked audiences. It is reported that the artifact based on face recognition, combined with multi-modal recognition, can distinguish cases that are difficult to distinguish by common biological recognition techniques such as identical twins, and more effectively avoid identity fraud caused by face forgery.

Recently, in the large-scale scientific and technological variety show "Intellectual Future", Alipay developed the face artifact for the first time, successfully found the target mission among 500 masked audiences. It is reported that the artifact based on face recognition, combined with multi-modal recognition, can distinguish cases that are difficult to distinguish by common biological recognition techniques such as identical twins, and more effectively avoid identity fraud caused by face forgery.

In fact, biotechnology represented by face recognition has long been the darling of the financial industry. Multimodal biometrics (ie, a combination of multiple biometrics) can help reduce the risk of financial fraud and ensure that user information or data is safer than a single biometric. Today, Xiaobian introduces the current status of mainstream biometrics in the financial industry.

Face recognition: wide application and controversy

With the development of the times, the application of financial market scenarios has become more diversified and complicated, and pure financial outlets have been unable to meet customer needs. Financial services anywhere, anytime, and on demand can be invincible. In order to achieve cost reduction and efficiency increase, emerging Internet financial institutions and traditional financial institutions have deployed face recognition technology to ensure the accurate and fast implementation of user identity authentication under various financial service channels, and improve the security and timeliness of business processing.

At present, face recognition technology has been applied to account registration, remote account opening, face payment, loans, etc., and most of these services require real-name authentication. In terms of real-name authentication, traditional authentication methods such as artificial naked eye judgment, SMS verification, and binding of bank cards have problems such as low accuracy, poor customer experience, and high cost. The use of face recognition technology is undoubtedly a timely rain.

The application of new technologies such as mobile Internet and face recognition makes financial services naturally integrated into the user's economic ecosystem and life circle, but one thing to be aware of is how to effectively solve photos at both ends of the virtual world. Video and other anti-counterfeiting questionsThere are still some difficulties in the problem. CCTV has exposed the vulnerability of face recognition and reminded the users to pay attention to prevention. The State Administration of Markets also pointed out that the smart door with the face recognition function is locked in a high security risk.

There are three important parameters in the face recognition algorithm, namely the rejection rate, the false recognition rate, and the recognition rate. Face recognition technology-related enterprises continue to iteratively upgrade the algorithm, the purpose of which is to ensure that the recognition rate is faster, the pursuit of false rejection rate, false positive rate or higher safety factor. It is believed that with the increasing maturity of technologies for improving the security and accuracy of face recognition, face recognition will play a greater role in the financial industry.

Fingerprint recognition: The technology is mature, the golden period has passed

The current face recognition is hot, in contrast, the fingerprint recognition is slightly dazed. Fingerprint recognition technology entered the financial industry on a large scale as early as the 1990s. After years of accumulation, it has matured. By logging in to the operating system equipped with fingerprint recognition technology, the operator's identity can be more accurately confirmed, and the phenomenon of indiscriminate authorization and unauthorized authorization can be avoided, and the risk of being invaded can be reduced, and the responsibility can be traced even in the event of an accident.

At present, financial institutions basically perform human face verification verification by face network verification. Based on the new generation ID card with fingerprint information, the fingerprint information collected by the user and the fingerprint information stored in the ID card can be carried out. The comparison ensures the identity of the holder and can also be used for remote identification of users, identity authentication at the time of account opening, and the like. In addition to the banking operating system, mobile terminals such as mobile phones are also commonly equipped with a fingerprint module, and users can conduct business through mobile banking.

The weakness of fingerprint recognition technology is that fingerprints are left on the collection device when fingerprints are used, which is easy to be copied; some special groups have fewer fingerprint features and are difficult to image; they have requirements for the humidity and cleanliness of the fingers. Fingerprint wear is also unrecognizable.

Iris recognition and finger vein recognition: quite potential, bursting time is needed

Iris recognition and finger vein recognition are two technologies as a rising star, with high recognition accuracy, scalability and Security, but because it is costly and not convenient enough, it has not been widely used in the financial sector. At present, iris recognition is mainly applied to scenes with high security requirements such as access control management and escort management. The finger vein identification technology has also been applied in some bank self-service devices.

In recent years, the irisThe technology of identification and fingerprint recognition continues to develop and popularize, and user acceptance has also increased. Many mobile phone manufacturers are also working to add iris recognition technology to mobile devices to provide more secure encryption protection. Therefore, the potential of these two technologies should not be underestimated.

Conclusion: In recent years, the deep integration of finance and technology has been accelerating. In different application scenarios, the biometric technology required is not the same. A single biometric identification is difficult to meet the needs of future diversification, and multimodal biometrics will become a future trend in financial technology.