Google’s latest Pixel 10 Pro and XL models feature a groundbreaking camera system that harnesses generative AI to clean up distorted images, particularly those taken with digital zoom. This technology, dubbed “Pro Res Zoom,” aims to improve low-quality photos by infusing the camera itself with AI.
The new system uses a latent diffusion model to process images, which helps reduce artifacts and produces more accurate results, especially when used in extreme conditions like 100x digital zoom. Pro Res Zoom’s processing is done entirely on device after taking the photo, saving around 4-5 seconds of runtime from initial development time.
However, this innovation raises questions about what constitutes a “photo” and whether a camera that uses AI for image enhancement still qualifies as a traditional camera. Google has taken steps to address concerns by tagging photos with content credentials, indicating when they’ve been edited or processed using AI tools.
Is a camera that leverages AI for image improvement still just a camera? That’s a question Google’s product manager Isaac Reynolds can’t help but ponder. He acknowledges the need for education and awareness about AI-generated images, which could have significant implications for our institutions and individuals.
Source: https://www.theverge.com/hands-on/761817/google-pixel-10-pro-res-zoom-ai