hao-qiang / AM-Softmax Public

Notifications You must be signed in to change notification settings
Fork 11
Star 26

This is an AM-Softmax tutorial and keras implementation.

26 stars 11 forks Branches Tags Activity

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
AM-Softmax.ipynb		AM-Softmax.ipynb
README.md		README.md

Repository files navigation

AM-Softmax

this is a AM-Softmax tutorial and keras implement. $$ \begin{aligned} L_ams & = -\frac{1}{n}\sum_{i=1}^{n}{\log{\frac{e^{s(\cos\theta_{y_i}-m)}}{e^{s(\cos\theta_{y_i}-m)}+\sum_{j=1,j\neq y_i}^{c}{e^{s\cos\theta_j}}}}} \ & = -\frac{1}{n}\sum_{i=1}^{n}{\log{\frac{e^{s(w_{y_i}^T x_i -m)}}{e^{s(w_{y_i}^T x_i-m)}+\sum_{j=1,j\neq y_i}^{c}{e^{sw_j^T x_i}}}}} \end{aligned} $$