NYU-CV-Fall-2018

Assignment 2: Traffic sign competition

Requirements

1. Install PyTorch from http://pytorch.org

2. Run the following command to install additional dependencies

pip install -r requirements.txt

Training and validating your model

Run the script main.py to train your model.

Modify main.py, model.py and data.py for your assignment, with an aim to make the validation score better.

• By default the images are loaded and resized to 32x32 pixels and normalized to zero-mean and standard deviation of 1. See data.py for the data_transforms.

• By default a validation set is split for you from the training set and put in [datadir]/val_images. See data.py on how this is done.

Evaluating your model on the test set

As the model trains, model checkpoints are saved to files such as model_x.pth to the current working directory. You can take one of the checkpoints and run:

python evaluate.py --data [data_dir] --model [model_file]

That generates a file gtsrb_kaggle.csv that you can upload to the private kaggle competition https://www.kaggle.com/c/nyu-cv-fall-2018/ to get onto the leaderboard.