Recurrent age estimation

Open Webpage

Huiying Zhang, Xin Geng, Yu Zhang, Fanyong Cheng

Published: 2019, Last Modified: 21 Jul 2025Pattern Recognit. Lett. 2019EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Age estimation singly based on face image features does not achieve good results.•Recurrent age estimation (RAE) is introduced to age estimation.•RAE can learn personalized aging patterns by combining CNNs with LSTM.•RAE can process variable-length face images.•The extend label distribution learning (LDL) is employed to solve the issue of data shortage.