ADVANCE.AI participated in the year selection of high-quality enterprises and industry representatives of the 2019 Shu Ju Yuan· Golden Ape Prize, which initiated by China's authoritative big data media Shu Ju Yuan. ADVANCE.AI's independent developed solution for Digital Identity Verification (DIV) is selected in the Golden Ape Prize·2019 China Big Data Product Ranking List.
ADVANCE.AI's DIV solution effectively verifies the user identities and blocks fraudulent activities via computer vision technologies, including OCR, ID Recognition, Liveness Detection, Face Comparison, Risky Face Search, etc.
△ The workflow of ADVANCE.AI's DIV solution
ADVANCE.AI's DIV solution can help businesses to reduce the cost of the labour force. For example, when the DIV solution is not being used, customers need to go to the financial institution in-person to perform the verification process. Correspondingly, the financial institution needs to designate workers to perform the verification process as well. The whole process is tedious, time-consuming and costly, while the DIV solution can solve these issues only by using mobile equipment.
With the fast development of the digital bank, e-commerce and all other online businesses, regardless of whether it is a financial institute or an enterprise, a fast, efficient and cost-effective way to perform the identity verification process becomes necessary. ADVANCE.AI's DIV solution, which uses computer vision and anti-fraud technology, not only smooths the whole identity verification process that can be completed within 5 minutes but also intellectualizes the entire system and prevents fraudsters from using fake ID card or intentionally using falsely.
Use scenarios
The most common financial fraud is that fraudsters try to attach their own photos on someone's ID cards to cheat the verification process. Due to the different standard on the face recognition database for various counties and the inadequacy of the official face database, it is hard to detect the fraudsters using this way to cheat the verification process.
However, ADVANCE.AI's DIV solution can detect this fraudulent effectively. First, the applicant's ID card information is identified by OCR technology, and then, Liveness Detection technology needs to scan the applicant's actual face image during the verification process. Finally, the DIV solution will compare the liveness face with the ID card image to judge whether the applicant adopts the real and legal identity.
ADVANCE.AI's DIV solution can store the image to the face image database during the verification process. When the applicant tends to use a different ID card in the same or other financial institutes, the solution will compare all the applicant's face images to determine whether this application has fraudulent intention.
Product features
1) OCR
a. Recognize ID card information automatically; minimize human errors; improve job efficiency.
2) Liveness Detection
a. To ensure the person being verified is actually a live person. It is based on the huge movement imaging and repeating data algorithm to eliminate fraudsters and fake ID card. Liveness Detection includes movement detection and re-shooting detection.
3) Face Comparison
a. This service is specially developed for the DIV solution. The similarity rate of the pictures will be inputted during the verification process to increase the success rate and lower the failure rate caused by the damage of the ID card or other unexpected situations. This service tested using a massive amount of data imaging sample model. The accuracy rate is the best among similar products in the Southeast Asia market.
4) Face Search
a. With ADVANCE.AI's high-resolution face recognition technology, Face Search can be used to dig out the identity fraud in trading transaction records effectively and avoid subsequent fraud.
b. This service is the best in the area where the government identity information is not available, also in those financial institutes of the countries that the related information is not properly collected.
5) Risky Faces
a. Identify the risk from the face image data, determine fraudulent risk, provide fraudulent risk alert in real-time.
b. Depend on the accuracy of the database information in the ADVANCE Guardian anti-fraud detection services. The database updates continuously to provide more reliable information to ensure ADVANCE.AI's fraudulent database uniqueness and reliability.
Technology advantage
ADVANCE.AI's DIV solution uses advanced face recognition calculation, which has an accuracy rate as high as 96%, the error ratio lower than 1 of 100,000. The accuracy rate is the highest among similar products in Southeast Asia. This calculation method can provide the highest accuracy rate even the imaging conditions are not at the peak—for instance, weak lighting condition; headwear effect; low-quality image, etc.
Thanks to the current colossal data accumulation, sophisticated data processing, clear labelling and tremendous DIV application data testing. ADVANCE.AI's face recognition technology has a hit rate of over 99% searching in a million application scenarios.
Clients
ADVANCE.AI has served over 300 business clients in Indonesia, the Philippines, Vietnam, and other Southeast Asian countries spread out over the banking, fintech, telecommunications, travel and e-commerce industries with a monthly API visit of over 40 million.
Market value
There are many criminal groups that take advantage of the incomplete data information infrastructure in Southeast Asia countries to perform illegal, fraudulent activities that are disturbing the financial market order. With advanced computer vision and data mining technologies, ADVANCE.AI has successfully helped many financial companies in Southeast Asia countries block over 100 thousand fraudulent applications and help clients avoid tremendous financial losses. So far, ADVANCE.AI clouds received millions of different product usages daily and accumulated over 100 million data images in the last three years. It will help ADVANCE.AI to devote serving more reliable, more efficient anti-fraud solutions for business clients.
