Fairsense-AI utilizes LLMs for generating detailed analyses of textual content, detecting biases, and providing insights on AI governance topics.
Fairsense-AI uses Blip for generating descriptive captions of images. This aids in understanding visual content and assessing it for biases or sensitive elements.
Fairsense-AI employs Tesseract OCR to extract text from images, allowing analysis of textual content embedded within images.
Transformers (Hugging Face) and PyTorch power the underlying models, ensuring robust NLP and deep learning functionalities.
Plotly is used for creating interactive charts in the AI Safety Risks Dashboard, providing engaging and informative data visualizations.
Gradio offers a clean, user-friendly UI for interacting with the Fairsense-AI platform.
Shaina Raza PhD, Marcelo Lotif, Tahniat Khan, Mukund Sayeeganesh Chettiar, Aravind Narayanan, Matin Yousefabadi
For queries, contact: shaina.raza@vectorinstitute.ai