Posts

Her work explores how AI systems can reason directly in dynamic 3D environments rather than only through text, including methods that extend 4D capture with generative models.

The award will allow Chen to create a scientific framework that allows machines to learn directly from passive observations and active interactions with the physical environment.

This year’s event was held from April 10-11 and Brown CS student Eashan Iyer, who expects to concentrate in Mathematical Physics and Computer Science, was part of a cross-university team that placed second in the competition.

The waves are calling, the blue expanse with its endless opportunities to tinker with oil coolers and hose clamps and heat exchangers, to pull up alongside other sailors at the close of day and be social or not. John is weighing anchor: he’s retiring from Brown after 42 years of teaching, with 32 spent at Brown CS.

The AAAS has elected Brown CS PhD alum Danfeng (Daphne) Yao (now Professor of CS, Elizabeth and James E. Turner Jr. '56 Faculty Fellow, and CACI Faculty Fellow at Virginia Tech) to the rank of Fellow.

Brown CS is happy to announce that with the anticipated approval of Brown’s Corporation and effective as of July 1, 2026, Robert Y. Lewis has been promoted to Associate Teaching Professor.

Brown CS is happy to announce that with the anticipated approval of Brown’s Corporation and effective as of July 1, 2026, Lorenzo De Stefani has been promoted to Associate Teaching Professor.

Imagine watching a concert not from a fixed camera angle, but from any angle. The catch? Volumetric video is incredibly hard to store and stream, and you can have the most photorealistic 4D scene in the world, but if you can't get it to a viewer efficiently, it's stuck in a lab. Our work, PackUV, tackles exactly this problem.

In this paper, we leverage a 3D Geometric Foundation Model to build a self-supervised pipeline that evaluates 3D consistency in AI-generated videos. By integrating our video generation model with reinforcement learning, we are able to generate highly 3D-coherent and realistic videos. This approach significantly reduces morphing, flickering, and artifacts, outperforming current state-of-the-art methods.

No stranger to controversy, AI pioneer Yann LeCun told a capacity crowd that large language models are not the future of AI and that a new approach is needed if machines are to achieve human-like intelligence.