Federal Project Grant Award Summary

Virginia Polytechnic Institute & State University (Virginia Tech) received a $325,054 Project Grant award from the National Science Foundation's (NSF) Division of Computing and Communication Foundations under the Computer and Information Science and Engineering (CISE) program (CFDA 47.070), effective August 1, 2026 through July 31, 2031. This CAREER award supports research toward generative cinematography with controllable video synthesis, addressing the challenge of making high-quality video creation more accessible and controllable through artificial intelligence.

The award funds development of training-free methods to unlock controllable capabilities within pre-trained video generation models, enabling users with director-like control over visual content, motion, and camera dynamics without requiring additional model training or curated datasets. Research is organized across three interconnected thrusts: establishing foundational understanding of text-to-video model mechanisms; developing training-free control methods for content, motion, and camera viewpoints; and integrating capabilities into a multi-agent collaboration framework with autonomous reasoning and self-repair capabilities. The project will produce open-source tools, datasets, and interactive demonstrations while advancing theoretical understanding of latent representations in video diffusion models. Complementary education and outreach activities target audiences from elementary school students to creative industry professionals to broaden understanding of generative artificial intelligence.