CNN_AidedDiagnosisTool

Software for Diagnosis based on Convolutional Neural Network. More informations available in the file: Project Description.

Preliminary rules

1. Note that independence and autonomy in solving the project assignment is a matter of evaluation. Please avoid to ask questions about implementation issues. This kind of problems must be solved by yourselves, you can ask but such questions are kept into consideration for the final evaluation or even ignored.
2. All the assignment are mandatory, each variation or change must be previously discussed.
3. For the final debate you must provide: (following section for specifications)
4. PDF report
5. Set of slides used during discussion
6. Code of the complete pipeline
7. During oral discussion any kind of questions related to the course topics might be asked.

Required documentation

Must be submitted 3 days before exam date

• Report PDF: such document must be organized with the following format:
  1. Description of the assignment of the project and the focus of your work;
  2. Theory about CNNs focused on the application outline, providing references;
  3. Detailed description of the method you used (describe most salient aspects of your code too);
  4. Detailed description of the results (provide graphs, tables, etc.);
  5. Results discussion;
  6. Future development;
• Slide: must be used during oral discussion. Follow the structure of your report.
• Code: all the code you used must be provided. Carefully comment it and keep it well organized. Such criteria are matter of evaluation.

All the documentation will be available in the documents folder.

Environment

Before launching the project is important to install the proper package (make use of Virtual Environment) by executing the command:

pip3 install -r requirements.txt

The folder with all the dataset must be named:

~/dataset

and must be created in the project root.

Built With

• Python3 - The development language
• TensorFlow - Machine Learning framework
• PyCharm - Python IDE by JetBrains

Authors

• Jacopo Maggio - JackMGit
• Jacopo Nasi - Jacopx
• Sofia Ostellino - aSofworkOrange

License

The property of this project belongs to the Polytechnic of Turin since this project isn't delivered yet. After that it will be published under not-already defined license.