Classifying 5 types of Flowers using Deep Learning.

Rose: 0.986	Dandelion: 0.93

The dataset is from Tensorflow's [Flowers Recognition]( http://download.tensorflow.org/example_images/flower_photos.tgz ). The goal is to classify five kinds of flowers (daisy, dandelion, roses, sunflowers, tulips) by raw image.

The dataset contains 3670 images of flowers. The pictures are divided into five classes: daisy, dandelion, roses, sunflowers, tulips. For each class there are about 700 photos.

1. Resize all the input images to 48x48.
2. 90% training samples && 10% validation samples.

python3 train.py

python3 predict.py <filename>

• Computing: Google Colab Tesla K80 GPU
• Python version: 3.6.6
• Using packages
  1. Keras (tensorflow.python.keras) for building models
  2. OpenCV (cv2) for processing images
  3. sikit-learn (sklearn) for train_test_split
  4. Install necessary modules with sudo pip3 install -r requirements.txt command.