1 paper
Yiran Wang, Oliver Turner, Viorela Ila
Existing Vision Foundation Model (VFM)-based iterative stereo pipelines under-exploit three information pathways: multi-scale backbone features are collapsed into single-level corr…