Color Classifier

Color Classifier is an assignment of Cloud Counselage. The assignment provided hands-on experience with tensorflow

Link to saved model

My approach

Initially I webscraped images from shutterstock, however while webscraping some images were corrupt and caused a bad prediction due to these images

To run the webscraping code

python webscrape.py

I created images using cv2 where I create random colors

Adjust the parameters in the code & execute image_maker

python image_maker.py

Then I trained a neural network model using tensorflow 2.0 and saved the model

My method explained in the notebook ColorClassifer.ipynb

I have trained using 600 images and validated on 150 images

Finally I wrote a script to predict Color of images in test dataset

Store the images in testing folder and execute image_predict

python image_predict.py

Name		Name	Last commit message	Last commit date
Latest commit History 9 Commits
testing		testing
training		training
validation		validation
ColorClassifier.ipynb		ColorClassifier.ipynb
Heurestic.ipynb		Heurestic.ipynb
Output.png		Output.png
README.md		README.md
image_maker.py		image_maker.py
image_predict.py		image_predict.py
webscrape.py		webscrape.py