HOME CHALLENGES DATASETS TASKS LEADERBOARDS SUBMIT RESULTS Sign in

Submission: CDPN_BOP19 (RGB-only)/HB

Download submission
Submission name
Submission time (UTC) Aug. 17, 2020, 1:24 p.m.
User fbw19
Task Model-based 6D localization of seen objects
Dataset HB
Training model type Default
Training image type Synthetic (only PBR images provided for BOP Challenge 2020 were used)
Description
Evaluation scores
AR:0.672
AR_MSPD:0.816
AR_MSSD:0.642
AR_VSD:0.557
average_time_per_image:0.280

Method: CDPN_BOP19 (RGB-only)

User fbw19
Publication CDPN: Coordinates-Based Disentangled Pose Network for Real-Time RGB-Based 6-DoF Object Pose Estimation
Implementation https://github.com/LZGMatrix/BOP19_CDPN_2019ICCV
Training image modalities RGB
Test image modalities RGB
Description

This method stays the same as Zhigang-CDPN-ICCV19 in BOP Challenge 2019. For LMO, HB, ICBIN and ITODD datasets, only PBR images provided for BOP Challenge 2020 were used during training. For YCBV, TUD and TLESS datasets, both PBR images provided for BOP Challenge 2020 and real images provided for BOP Challenge 2019 were used during training.
Computer specifications CPU: Intel i7-7700; GPU: GTX 1070