Submitted by Ekrem Alper Kesen.

Submission data

Full nameScanRefer + BRNet + DGCNN + Self-Attention + Cross-modal Attention (end-to-end)
DescriptionUses BRNet for object detection, self-attention for language embeddings, cross-modal attention for visual and textual embeddings, apply DGCNN to fused features
Input Data TypesUses XYZ coordinates,Uses Multiview Image Features,Uses Normal Vectors
Programming language(s)Python
HardwareGeForce RTX 2080 Ti
Source code or download URLhttps://github.com/yoonhachoe/ScanRefer-GAB
Submission creation date8 Jul, 2021
Last edited15 Jul, 2021

Localization

Unique Unique Multiple Multiple Overall Overall
acc@0.25IoUacc@0.5IoUacc@0.25IoUacc@0.5IoUacc@0.25IoUacc@0.5IoU
0.70160.52020.32330.19590.40810.2686