Planktos is an innovative image recognition model tailored for classifying various species of phytoplankton. Built using cutting-edge machine learning algorithms, this tool enables researchers and environmentalists to automate and accelerate the identification process, enhancing studies in marine biology and ecology.
- Customizable Model Training: Fine-tune the model parameters based on your specific dataset requirements.
- Efficient Prediction: Utilize the trained model to classify new images quickly and accurately.
- Flexible Data Preprocessing: Ready your dataset for training with a robust preprocessing script tailored for phytoplankton image data.
Clone the repository and install the necessary dependencies:
git clone https://github.com/cee8/planktos.git
cd planktos
pip install -r requirements.txt
Start the training process with the following command, which will also handle model serialization and label saving:
python3 metaflow/phytoplankton_flow.py run
To predict phytoplankton classes from new images:
python3 src/predict_image.py
Prepare your dataset by running the preprocessing script:
python3 src/data_loader.py
We welcome contributions to Planktos! For more information on how to help, please refer to the contributing guidelines.
This project is licensed under the MIT License - see the LICENSE file for details.
Feel free to raise any issues or suggestions on the GitHub page. Alternatively, you can contact me directly at https://www.linkedin.com/in/beaudoin-/.