On August 22nd, a big data finance forum with the theme of "shaping the data ecosystem, leading financial technology, and focusing on the future of data" opened in Shanghai. Zhou Shouye, Vice President of Minivision Technology, and over 500 decision-makers and professionals from well-known companies in the industry such as regulatory authorities, authoritative associations, and fintech companies jointly attended the forum. Focusing on a group of financial technology companies emerging in the new financial environment that promote financial service reform through technological innovation and improve financial service efficiency through disruptive innovation, we shared our insights on using Face as the main ID, combining artificial intelligence with big data, and creating an intelligent ecosystem to serve multiple financial scenarios.

As a core part and important support of financial technology, big data finance is manifested in multiple fields such as credit reporting, online lending, payment, comprehensive financial services, intelligent investment advisory, and internet insurance. The forum focuses on the application and innovation of big data in the financial field, especially in the era of financial technology. On the basis of revitalizing existing data resources, how to establish a more comprehensive data ecosystem and financial scenarios, and strengthen the construction of risk control and credit reporting systems. Adhering to policy hotspots and facing industry difficulties, we invite industry elites to provide intellectual support and cutting-edge analysis for strategic transformation and innovative development of enterprises.

Zhou Shouye, Vice President of Minivision Technology, mentioned in his speech that Xiaoshi Technology's foothold is to explore the intelligent ecosystem of AI+big data based on deep learning. In financial scenarios, the first door to guarding financial security is to intercept identity fraud, and currently the only relatively effective method is facial recognition. Since 2015, Minivision Technology has been investing in the development of image recognition based on machine learning technology. Starting from image recognition, carry out financial anti fraud work. At present, personal information and photos circulated and sold on the black market are not difficult to obtain. Loan fraudsters upload processed photos of others and disguise themselves as living people. These dynamic photos can be used to create micro expressions such as blinking as required, avoiding the steps of live body recognition and muddling through. Minivision Technology has launched silent living technology in the industry and has been certified by the Third Institute of Public Security. This technology can accurately and efficiently identify the fraudulent means of "fake face pretending to be real face". Without the need for users to cooperate with micro expressions and other actions, real-time analysis of facial images is directly carried out through the viewfinder to capture differences in the video sequence, thereby determining whether the operator is a real person and ruling out the possibility of photo attacks. Xiaoshi Technology currently has over ten million daily API calls, effectively blocking nearly ten million identity impersonation and fraudulent means, saving customers nearly 20 billion in economic losses. Minivision Technology will extract feature vectors from each face and label it with corresponding loan application labels, which can identify whether the user has multiple borrowing risks on multiple platforms based solely on the face. Similar application scenarios can also be extended to eliminate incidents such as group wool harvesting. Minivision also has in-depth research and involvement in digital marketing and smart risk control. Xiaoshi ultimately wants to provide customers with intelligent asset allocation solutions that are suitable for their attributes.

At the same time, Minivision provides risk control engine tools and whitelist services to screen user groups with a credit rating or above recognized by the risk control system for customers and conduct intelligent pricing. Designate different levels of whitelist pools for customers' C-end users to provide intelligent asset allocation service solutions. This service allows users to receive loans within minutes after facial recognition. Simultaneously maintaining the defect rate within 5% greatly improves customer experience. At present, the whitelist (consumer whitelist) service has been successfully applied in industries such as live streaming, travel, and e-commerce.