LetheSec/Fer2013-Facial-Emotion-Recognition-Pytorch
This method achieves SOTA single model accuracy of 73.70 % on FER2013 without using extra training data.
PythonMIT
Issues
- 1
Question about other model
#9 opened by VietPT3502 - 1
训练集与验证集ACC与LOSS异常
#8 opened by ZifengLiu98 - 2
train acc 比 test acc 低20% 是为什么呢?
#7 opened by xioapangzi - 0
- 2
mixup_data
#2 opened by jiantenggei - 0
您好,关于网络集成的疑惑
#5 opened by thgpddl - 8
作者您好,如何能达到73.7%的test精度
#3 opened by thgpddl - 1