CS231N_Project

Final project for CS231N Deep Learning for Computer Vision at Stanford University Authors:

Authors:

A. Running the code

Data Preprocessing

python data_process.py

Clone Repository

git clone git@github.com:jacobazoulay/CS231N_Project.git

Create Virtual Environment

$PYTHON_BIN = path/to/python/bin
virtualenv -p $PYTHON_BIN venv
source activate venv/bin/activate

Install requirements

pip install -r requirements.txt

Training/Testing a network

1. Set parameters in run.sh
2. Run: ./run.sh
3. Results are saved in './results' directory

Plotting loss after training

python tools/plot_loss.py path/to/results/trainlog.txt

Visualizing Predictions on Images

1. Run evaluation by setting train=0 in run.sh
2. Predictions are stored in path/to/results/Images