Submission: FCOS-CDPN-PBR/LM-O

Submission name
Submission time (UTC) July 23, 2022, 9:59 a.m.
User Yang-hai
Task Detection (BOP 2022)
Dataset LM-O
Training model type None
Training image type Synthetic (only PBR images provided for BOP Challenge 2020 were used)
Description
Evaluation scores
AP:0.570
AP50:0.900
AP75:0.628
AP_large:0.659
AP_medium:0.621
AP_small:0.160
AR1:0.607
AR10:0.650
AR100:0.659
AR_large:0.722
AR_medium:0.698
AR_small:0.381
average_time_per_image:0.040

Method: FCOS-CDPN-PBR

User Yang-hai
Publication CVPR-2019
Implementation https://github.com/LZGMatrix/BOP19_CDPN_2019ICCV/tree/bop2020
Training image modalities RGB
Test image modalities RGB
Description

As CDPN provides the trained detectors(FCOS+VoVNet 57) used in the 2020 6D Pose challenge, we test them to provide more comprehensive results for the detection track.

Computer specifications NVIDIA 3090