zyang-ur / SAT

SAT: 2D Semantics Assisted Training for 3D Visual Grounding, ICCV 2021 (Oral)

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Cannot reproduce the results in the paper

CurryYuan opened this issue · comments

I run the training code with command provided in README, the results are lower than that in the paper.

Accuracy SAT Reproduce
Nr3d 49.2 46.0
ScanRefer 53.8 51.4

Do you have any idea about where the gap comes from?

Hi, is there any command for ScanRefer? CAN not find in README. Thanks!