Advanced AI-Face detector
SyntheticeEye is developing free and reliable detectors for AI-generated content.
Aletheia is a machine learning model designed to detect AI-generated faces and distinguish them from real faces. This model can be used on our website. Aletheia has achieved high accuracy in detecting images from various face generators, but we are still working on improving this model further and making it more robust.
Argus is our detector for AI-generated images in general. This model aims to detect a broad spectrum of AI-generated images. Argus can be tested on our website
We are working on developing models, that are able to detect more types of content. These are a few detectors we want to develop in the future:
- Text Detection
- Audio Detection
- Video Detection
With our models, we want to provide everybody with reliable, free, and user-friendly tools that help them detect AI-Generated Content. Our models help verify content and detect misinformation. With the rapid advancements of generative AI, tools like ours become essential to preserve trust.
- assets/: Storage for project-related images.
- notebooks/: Jupyter notebooks for our models.
- runs/: Stores logs from notebooks. These are not included in the repo, to ensure a lightweight repository.
- scripts/: Contains primarily helper functions.
- state_dicts/: Here we store our models state_dicts temporarily. They are not included in the repo, due to their large file size.
- documentation: Contains model-specific requirements
- Aletheia: Requirements for Aletheia.
- Argus Requirements for Argus.
Test our models on our website at syntheticeye.dev
Use our models directly in the browser with our chrome extension