Amazon Rekognition improves face search accuracy with user vectors
Today, AWS launched a new capability that significantly improves face search accuracy by leveraging multiple face images of a user. Currently, Amazon Rekognition allows customers to search users represented by individual face vectors. Face vectors are mathematical representations of faces from images. Now, customers can create user vectors, which aggregate multiple face vectors of the same user. User vectors offer higher face search accuracy with more robust depictions, as they contain varying degrees of lighting, sharpness, pose, appearance, etc.
Today, AWS launched a new capability that significantly improves face search accuracy by leveraging multiple face images of a user. Currently, Amazon Rekognition allows customers to search users represented by individual face vectors. Face vectors are mathematical representations of faces from images. Now, customers can create user vectors, which aggregate multiple face vectors of the same user. User vectors offer higher face search accuracy with more robust depictions, as they contain varying degrees of lighting, sharpness, pose, appearance, etc.
Financial services, gig economy, telco, healthcare, photo sharing, media companies, and other customers use Amazon Rekognition Face Search for identity verification (IDV) during online onboarding, step-up authentication, and person search. For example, IDV providers use Face Search to determine whether a new user’s face matches an existing account or a known fraudulent offender. Customers can now improve accuracy of search by aggregating up to 100 face vectors into one user vector for each user. Compared to individual face vectors, user vectors offer stronger security and a better overall experience.
User vectors are now available in all commercial AWS regions supported by Amazon Rekognition Faces. To get started, please download the latest AWS SDK and visit our Amazon Rekognition Faces documentation.