This project is based on python, keras, sklearn ... Thus you have to setup an environment with the following libraries. I also provide a requirments.txt file containing all the necessary requisites and their prerequisites:
- Keras (2.2.4)
- Numpy (1.15.4)
- Pandas (0.24.2)
- pillow (5.4.1)
- matplotlib (3.0.3)
- scipy (1.1.0)
As this project is based on image recognition, you have to download the training data. As the required data is quite big, you will have to download it from the initial provider kaggle. Therefore, please install the Kaggle API and use the following command to download the data: kaggle competitions download -c humpback-whale-identification
After downloading the data, please extract the train.zip to the subfolder "data/train" in your project folder and extract the test.zip to the subfolder "data/test".
a