Skip to content

ck-kai/SelfTraining-6D

Repository files navigation

SelfTraining-6D

This repo provides for the implementation of the ECCV'22 paper:

Sim-to-Real 6D Object Pose Estimation via Iterative Self-training for Robotic Bin Picking arXiv

Requirement

  • Ubuntu 18.04, CUDA 10.2, Python >= 3.6
  • kaolin == 0.1.0
  • opencv-python == 4.5.4.58

Installation

Compile the knn module:

cd lib/knn
python setup.py install --user

Compile the ransac voting layer:

cd lib/ransac_voting
python setup.py install --user

Install kaolin

git clone https://github.com/NVIDIAGameWorks/kaolin.git
cd kaolin
git checkout v0.1
python setup.py develop

Dataset

Download our processed ROBI dataset from here and put them into 'SelfTraining-6D/data'

Virtual Training

Following object-posenet to train an object pose estimation model on our provided virtual data. Put the virtual model into 'SelfTraining-6D/virtual_models'. To skip this step, you can download our provided virtual model from here.

Sim-to-Real Training

self_training.py is the main file for sim-to-real self-training.

Example:

python self_training.py --dataset zigzag --nepoch 30 --iter 10

The intermediate data with pseudo labels will be stored into 'SelfTraining-6D/data'. The trained model will be stored into 'SelfTraining-6D/real_models'

Evaluation

Example:

python evaluate.py --obj_name zigzag --testing_mode st --testing_iter 5

Citation

If you find this repo helpful, please consider citing:

@InProceedings{chen_2022_sim,
  title     = {Sim-to-Real 6D Object Pose Estimation via Iterative Self-training for Robotic Bin Picking},
  author    = {Chen, Kai and Cao, Rui and James, Stephen and Li, Yichuan and Liu, Yun-Hui and Abbeel, Pieter and Dou, Qi},
  booktitle = {European Conference on Computer Vision (ECCV)},
  month     = {October},
  year      = {2022}
}

Any questions, please feel free to contact Kai Chen (kaichen@cse.cuhk.edu.hk).

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Languages