1 paper
Huimin Wu, Kwang-Ting Cheng, Stephen Lin +1
This paper presents an investigation of vision transformer learning for multi-view geometry tasks, such as optical flow estimation, by fine-tuning video foundation models. Unlike p…