hyBlue/STL-10-Classification-using-Keras

STL-10 Classification using Keras

Dataset

Use 8,000 images in test data as traing. Split training data into 1,000 and 4,000 images each for validation and test.

Environment

Google Colaboratory(code)

Model

Result

Top 1 accuracy: 78.625%
#Parameters: 2,571,466

Reference

https://appliedmachinelearning.blog/2018/03/24/achieving-90-accuracy-in-object-recognition-task-on-cifar-10-dataset-with-keras-convolutional-neural-networks/