Rundi Wu

Papers

Total Citations

H-Index

About

Rundi Wu is a rising researcher in computer vision and graphics, whose work pushes the boundaries of dynamic scene understanding and novel view synthesis. Her most-cited paper, "Generative Camera Dolly: Extreme Monocular Dynamic Novel View Synthesis" (2024, 9 citations), introduces a groundbreaking framework that generates photorealistic, temporally consistent videos from a single monocular input, enabling dramatic camera movements around dynamic scenes. This work addresses a long-standing challenge in 3D vision—synthesizing novel views of moving objects with large viewpoint changes—by leveraging generative models to hallucinate unseen content while preserving motion coherence. Wu’s contributions are particularly notable for their potential to democratize high-quality video production and virtual cinematography, requiring only a single camera. Her research has already garnered attention for its practical implications in augmented reality, filmmaking, and robotics. With a focus on bridging generative AI and geometric understanding, Wu is establishing herself as a key voice in the next wave of neural rendering, where static assumptions give way to dynamic, real-world complexity.