Pick a Pic: A Platform for Human Feedback on AI-Generated Images

Pick a Pic is an open-source application that empowers researchers and enthusiasts alike to contribute to the development of AI-generated images through human feedback. By leveraging Stable Diffusion 2.1 and ProtogenX3.4 models, users can generate images from text and rank them, providing valuable insights into the text-to-image dataset.

Mission-Driven Research

All code, data, and models resulting from this research project are openly available to the public, aligning with our mission to build the largest open-sourced human-feedback repository. By facilitating collaboration and transparency, we foster a community-driven approach to AI research.

Participation Requirements

To participate in the project, users must authenticate via Google, ensuring that all collected feedback is genuine and reliable. This commitment to data quality allows us to create a comprehensive dataset that accurately reflects human perception and understanding of AI-generated images.

Data Release and Updates

The collected data will be periodically updated and released into the public domain, providing a transparent and collaborative research environment. This open-access approach enables researchers to build upon existing knowledge and accelerate breakthroughs in AI development.

Generous Support from Google TRC Program

We are grateful for the generous support from the Google TRC Program, which further solidifies our commitment to academic excellence and open collaboration. Through this partnership, we can continue to drive innovation and push the boundaries of what is possible with AI-generated images.