Models trained to predict dog and cat breeds
This repository holds pipelines and models that were trained on dog and cat breed images and datasets to produce models for predicting the breed of a cat or dog from an image.
- image-classification
- python
- neural-networks
- pytorch
- artificial-intelligence
- machine-learning
- keras
Please see the academic poster for research notes and further information.
This project was written in python using Jupyter Notebook/Google Colab. To open any of the notebooks, please download and open in a notebook of choice.
To clone this repository, please clone as usual and then run "git lfs install" and "git lfs pull"
To test the end-to-end pipeline that was implemented with the final models, please use the notebook 'pipline_testing.ipynb'.
Press 'run all' to run this notebook. The notebook will prompt you to upload images that will be used to demonstrate the functionality of the developed object detection and classification models.
Additionally, the final models that are used by this notebook can be found in the 'models' folder.
** For reference only** The notebooks that were used to conduct model experimentation are provided in the 'Model Experiments' folder. These notebooks are used to create the models are present here. These notebooks can be executed through a 'run all' command, however the training duration is extremely long, so this is not recommended.