Skip to content

Files

Latest commit

5bb3ead · Jun 28, 2018

History

History
7 lines (4 loc) · 220 Bytes

README.md

File metadata and controls

7 lines (4 loc) · 220 Bytes

QSTP_Image_recognition

project on fashion MNIST

CNN trained on FashionMNIST dataset.

It is rather simple, 2 Conv layers with with sigmoid activation and maxpooling. The final accuracy of model on test set is 87.63%