Highest validation accuracy achieved: 79.1%
Highest deployment accuracy achieved: 38.0%
-
Specify a .csv file to use as the metadata of the images.
-
Specify a folder containing all of your images.
-
If using a custom image folder, simply resize the images to the correct resolution by clicking the 'Resize' button.
-
If you'd like to train the models on your own metadata and images, you can use the 'Train' button. Note: this may take a long time to complete, and will OVERWRITE existing models. Progress can be tracked in the accompanied command window.
-
You can choose between using the pre-trained default models, or your own trained models.
-
Select which model you would like to use.
-
Click 'Predict' and the resulting predictions will be shown in a new window.
-
Clone the repository locally.
-
Ensure Python 3 is installed, as well as all dependencies, which can be found by the command
$ pip check.
- Run the included shortcut "Cancer Predictor".
Please contact me or open an issue if you have any issues!