Introducing the Tufts Face Database
A group of Tufts researchers across the Schools of Arts & Sciences and Engineering have created an image database using multiple modalities. Incorporating photograph images, thermal images, near infrared images, recorded video, computerized facial sketches, and 3D images, the team has amassed a collection of more than 10,000 images from 113 individual volunteers. The volunteers represent a diverse set of ethnicities, genders, and countries of origin.
In their paper titled, "A comprehensive database for benchmarking imaging systems," the group presents the Tufts Face Database: the most comprehensive, large-scale, multi-modality face dataset available to the public. The team sees the public availablilty of this database as an asset to other researchers in further development of face recognition methods across multimodal images. The findings were published in the journal IEEE Transactions on Pattern Analysis and Machine Intelligence.
The research group is led by Dr. Karen Panetta of the Department of Electrical and Computer Engineering (pictured with doctoral candidate Qianwen Wan), and includes researchers from the ECE Department and the Department of Psychology.