The coffee-ai from skeltonmod

coffee-ai's Introduction

This is how you should structure your data models.

dataset/
    train/
        unripe/
            unripe_image1.jpg
            unripe_image2.jpg
            ...
        semi-ripe/
            semi_ripe_image1.jpg
            semi_ripe_image2.jpg
            ...
        ripe/
            ripe_image1.jpg
            ripe_image2.jpg
            ...
        overripe/
            overripe_image1.jpg
            overripe_image2.jpg
            ...
    test/
        unripe/
            unripe_image1.jpg
            unripe_image2.jpg
            ...
        semi-ripe/
            semi_ripe_image1.jpg
            semi_ripe_image2.jpg
            ...
        ripe/
            ripe_image1.jpg
            ripe_image2.jpg
            ...
        overripe/
            overripe_image1.jpg
            overripe_image2.jpg
            ...

This works best with NVIDIA GPUs that supports CUDA.

Installation

Install the required packages
```
pip install -r requirements.txt
```
Install CUDA and cuDNN if you have an NVIDIA GPU
- CUDA
- cuDNN

Usage

Inorder to get accurate results, run trainer.py and strucutre your dataset as shown above
To run the model on your own dataset, change the path in trainer.py to your dataset path

To use the trained model replace the model variable as below

  from keras.models import load_model

  # Load the saved model from the .h5 file
  model = load_model('path/to/model.h5')

  # Use the loaded model to make predictions
  predictions = model.predict(data)

Setup

Open a terminal inside the directory and run the python venv module to create a virtual environment
```
python -m venv venv
```
- Activate the virtual environment
  - Windows
```
venv\Scripts\activate
```
  - Linux
```
source venv/bin/activate
```
Install the required packages
```
pip install -r requirements.txt
```
Install CUDA and cuDNN if you have an NVIDIA GPU (OPTIONAL)
Run the trainer.py file
```
cd trainer && python trainer_gui.py
```
After training the models you can run the main file
```
cd .. && python main.py
```

Recommend Projects