Fusion Detector

_{2024/2/12 • Project Introduction}

This is my graduation design work, it's Adversarial Examples Detector based on Multi-Model Feature Fusion in essence.

Project layout

Its more code-demanding than I initially thought, thus a decent project layout is required. The project is divided into several directories, each of which contains files differ from others':

• data: Pretrained models on a certain dataset and the dataset itself.
• metadata: Model structure and summary on parameters and output shapes.
• src: Source code of core logic and fundamentals.
• tools: Tools that help packing sources for deployment, generating adversarial datasets, etc.

You may notice that some of the directories (e.g. the data directory) is missing from GitHub. Despite GitHub's file size limit, data files shall not be uploaded to git repositories.

Since the project is still under development, you can email me ([email protected]) to get data files. Once the project is complete, those data files will be uploaded on a cloud storage platform.