The below project involves the learning of building a convolutional deep neural network to classify user-supplied images. Given an image of a dog, the algorithm will identify an estimate of the canine’s breed. If supplied an image of a human, the code will identify the resembling dog breed.
-
Clone the repository and navigate to the downloaded folder.
-
Download the dog dataset, human dataset and VGG-16 bottleneck features. Unzip to
path/to/dog-project/dogImages
,path/to/dog-project/lfw
andpath/to/dog-project/bottleneck_features
respectively. -
Obtain the necessary Python packages using "pip install -r /path/to/requirements/requirements-gpu.txt" if using a CUDA based GPU or "pip install -r /path/to/requirements/requirements.txt" for a CPU execution.
-
Open the notebook (requires Annaconda3 to be installed locally) and follow the instructions.
jupyter notebook dog_app.ipynb
This project was built on NVIDIA GTX960, if GPU is a limitation please use an AWS EC2 instance.