DreamDrone: Text-to-Image Diffusion Models Are Zero-Shot Perpetual View Generators.

Hanyang Kong, Dongze Lian, Michael Bi Mi, Xinchao Wang

Browse the full ECCV paper archive.