Skip to content

Latest commit

 

History

History
38 lines (25 loc) · 1.51 KB

README.md

File metadata and controls

38 lines (25 loc) · 1.51 KB
Raw

Object recognition and computer vision 2021/2022

Assignment 3: Image classification

Requirements

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

  2. Run the following command to install additional dependencies

pip install -r requirements.txt

Dataset

We will be using a dataset containing 200 different classes of birds adapted from the CUB-200-2011 dataset. Download the training/validation/test images from here. The test image labels are not provided.

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 64x64 pixels and normalized to zero-mean and standard deviation of 1. See data.py for the data_transforms.

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 kaggle.csv that you can upload to the private kaggle competition website.

Acknowledgments

Adapted from Rob Fergus and Soumith Chintala https://github.com/soumith/traffic-sign-detection-homework.
Adaptation done by Gul Varol: https://github.com/gulvarol