Method: HccePose (Faster-RCNN)

User SEU_WYL
Publication
Implementation
Training image modalities RGB
Test image modalities RGB
Description

Submitted to: BOP Challenge 2023

Training data: real + provided PBR

Training Datasets of LM-O are real and PBR-based datasets.

Used 3D models: default

Notes:

Setting: One network (ResNet34) per object was trained (without Refinement)

2D Bounding Box: Faster-RCNN (ZebraPose)

Batch Size: 24

Epoch: 38k

Learning Rate: 0.0002

Computer specifications

Public submissions

Date Submission name Dataset
2023-10-08 07:00 - LM-O