CIFAR-10 Prediction In Keras

'keras_cifar10_prediction.ipynb' predict a webcam photo by a neaural network that is trained by CIFAR-10 datas.

'keras_cifar10_prediction.ipynb' visualizes CIFAR-10 image files label prediction results by a ResNet model or a CNN model. A fail prediction is displayed red text. left side text is a fail prediction, right side is a correct label.

How to use

1. clone this repository using git:

git clone https://github.com/uchidama/CIFAR10-Prediction-In-Keras.git

2. cd to the folder and run jupyter notebook:

cd CIFAR10-Prediction-In-Keras
jupyter notebook

3. click 'keras_cifar10_prediction.ipynb' and run.

Installation

This software using Keras.

• Keras installation instructions.

If you want to run without to think keras and backend deeplearning frameworks, enter this command.

pip install tensorflow==1.4.1
pip install keras==2.1.2

Then, enter this command to install other using python packages.

pip install -r requirements.txt

Model Files

• cifar10_ResNet20v1_model.092.h5 is generated by cifar10_resnet.py in keras examples.
• keras_cifar10_trained_model.h5 is generated by cifar10_cnn.py in keras examples.

License

MIT