The HyperResNet model for color classification problem
colorClassification
The datasets have contained about 80 images for trainset datasets for whole color classes and 90 images for the test set. colors which are prepared for this application is y yellow, black, white, green, red, orange, blue a and violet. In this implementation, basic colors are preferred for classification. and created a dataset containing images of these basic colors. ( Link to dataset )
To train the model execute:
python train.py
To test the model execute:
python test.py -m ./link to the model/model.pth -i ./test/red.png -n 9