Fetching the paper…

Multi-View Large Reconstruction Model via Geometry-Aware Positional Encoding and Attention · Around