Get 3D scene from multi-view stereo without camera and pose data by pointmap representation DUSt3R
Get 3D scene from multi-view stereo without camera and pose data by pointmap representation DUSt3R
DUSt3R: Geometric 3D Vision Made Easy
arXiv paper abstract https://arxiv.org/abs/2312.14132
arXiv PDF paper https://arxiv.org/pdf/2312.14132.pdf
Project page https://dust3r.europe.naverlabs.com
Multi-view stereo reconstruction (MVS) … requires … estimate the camera parameters e.g. intrinsic and extrinsic parameters … introduce DUSt3R … without prior information about camera calibration nor viewpoint poses.
… cast the pairwise reconstruction problem as a regression of pointmaps, relaxing the hard constraints of usual projective camera models … this … unifies the monocular and binocular reconstruction cases.
… where more than two images … propose a simple yet effective global alignment strategy that expresses all pairwise pointmaps in a common reference frame.
… base … network architecture on standard Transformer encoders and decoders, allowing us to leverage powerful pretrained models.
… formulation directly provides a 3D model of the scene … depth … pixel matches, relative and absolute camera.
… DUSt3R can unify various 3D vision tasks and set new SoTAs on monocular/multi-view depth estimation as well as relative pose estimation …
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
LinkedIn https://www.linkedin.com/in/morris-lee-47877b7b
