Researchers at the University of Maryland have developed a groundbreaking technique that uses artificial intelligence to reconstruct discernible 3D environments from the reflections in the eyes of individuals. This approach is based on Neural Radiance Fields (NeRF), a technology capable of creating 3D scenes from 2D images. The technique involved capturing high-resolution images of a person looking into a camera and then isolating the reflections before determining the person's viewing direction. The team’s early work offers an intriguing view into a future where a series of simple images could reveal comprehensive details about one’s surroundings. Despite the considerable progress made, the researchers were confronted with several hindrances including inherent noise from the cornea, which complicates the separation of the reflected light from complex iris textures. To overcome these challenges, the team introduced several workarounds, such as cornea pose optimization and iris texture decomposition during training. Furthermore, a machine-learning technique known as radial texture regularization loss was used to isolate and enhance the reflected scenery. Yet, the technology is still far from practical application as it needs a controlled environment and struggles to accurately depict scenes in real-world use.