Submitted by Yiming Zhang.

Submission data

Full nameM3DRef-CLIP
Publication titleMulti3DRefer: Grounding Text Description to Multiple 3D Objects
Publication authorsYiming Zhang, ZeMing Gong, Angel X. Chang
Publication venueICCV 2023
Publication URLhttps://arxiv.org/abs/2309.05251
Input Data TypesUses XYZ coordinates,Uses Multiview Image Features,Uses Normal Vectors
Programming language(s)Python
HardwareNVIDIA RTX A5000
Websitehttps://3dlg-hcvc.github.io/multi3drefer/
Source code or download URLhttps://github.com/3dlg-hcvc/M3DRef-CLIP
Submission creation date14 Mar, 2023
Last edited18 Sep, 2023

Localization

Unique Unique Multiple Multiple Overall Overall
acc@0.25IoUacc@0.5IoUacc@0.25IoUacc@0.5IoUacc@0.25IoUacc@0.5IoU
0.79800.70850.46920.38070.54330.4545