GENKI Dataset is an expanding image database that covers faces of different lighting conditions, geographic locations, personal identities, and ethnicities. The dataset consists of three parts: GENKI-R2009a, GENKI-4K, and GENKI-SZSL. GENKI-R2009a contains 11,159 images, and GENKI-4K contains 4,000 images, which are divided into two categories: "smiling" and "not smiling". The face size, posture, lighting changes, and rotation angles of each image are different, and it is mainly used for smile recognition.

The GENKI Dataset was released in 2009 by the University of California – Machine Concepts Laboratory.