Submission: ZebraPose/LM-O

Download submission
Submission name
Submission time (UTC) Jan. 30, 2022, 2:52 p.m.
User zebrapose
Task Model-based 6D localization of seen objects
Dataset LM-O
Training model type Default
Training image type Synthetic (only PBR images provided for BOP Challenge 2020 were used)
Description PBR-Only and RGB-Only accepted at CVPR 2022 preprint: https://arxiv.org/pdf/2203.09418.pdf code: https://github.com/suyz526/zebrapose
Evaluation scores
AR:0.718
AR_MSPD:0.880
AR_MSSD:0.721
AR_VSD:0.552
average_time_per_image:0.250

Method: ZebraPose

User zebrapose
Publication ZebraPose: Coarse to Fine Surface Encoding for 6DoF Object Pose Estimation, CVPR2022
Implementation https://github.com/suyz526/ZebraPose
Training image modalities RGB
Test image modalities RGB
Description

Arxiv: https://arxiv.org/abs/2203.09418

We fixed a bug which influence the inference stage of Tless. The bug has been fixed on 17.Oct.2022. This new score is calculated using the same checkpoints. The old score was 0.603.

The reported inference time included 2D detection time.

Computer specifications Intel(R) Xeon(R) E-2146G CPU @ 3.50GHz, Nvidia RTX2080Ti