chainer-dense-fusion

This is Chainer implementation of DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion.

Original PyTorch repository is j96w/DenseFusion.

Requirement

• CuPy
• Chainer
• ChainerCV
• OpenCV

Installation

We recommend to use Anacoda.

# Requirement installation
conda create -n dense-fusion python=3.6
source activate dense-fusion 
pip install opencv-python
pip install cupy

# Installation
git clone https://github.com/knorth55/chainer-dense-fusion.git
cd chainer-dense-fusion/
pip install -e .

Inference

cd examples/dense_fusion/
python demo.py --random

TODO

• YCB Video Dataset
  • [x] Add estimator inference script.
  • [x] Add refiner inference script.
  • [ ] Add training script.
  • [ ] Reproduce original accuracy.

LICENSE

MIT LICENSE