1 paper · 1 filter
Haoran Bai, Xiaoxu Chen, Canqian Yang +3
We present Vivid-VR, a DiT-based generative video restoration method built upon an advanced T2V foundation model, where ControlNet is leveraged to control the generation process, e…