Our Team

My research centers on multimodal generative modeling – especially diffusion = for coherent, consistent, and controllable text-to-image, image-to-image, and text-to-video/audio systems. I’m interested in representations that align visual, auditory, and linguistic signals and then enforce global coherence: identity and style preservation across shots, stable layouts and scene geometry, long-horizon temporal smoothness, and tight cross-modal synchronization. I want to build generative systems that understand and compose across modalities while staying consistent for reliable, controllable content creation.