Diff-Tracker: Text-to-Image Diffusion Models are Unsupervised Trackers.

Zhengbo Zhang, Li Xu, Duo Peng, Hossein Rahmani, Jun Liu

Browse the full ECCV paper archive.