Multi-threshold deep metric learning for facial expression recognition

Wenwu Yang, Jinyi Yu, Tuo Chen, Zhenguang Liu, Xun Wang, Jianbing Shen

Abstract: Highlights•Triplet loss learning is highly advantageous for facial expression recognition.•Multi-threshold effectively increases triplet loss learning for expression recognition.•The difficult threshold validation for triplet loss learning is avoided.•Fast convergence in triplet loss learning is achieved.