Skip to content

lolrudy/LaPose

Folders and files

NameName
Last commit message
Last commit date

Latest commit

081b37a · Dec 23, 2024

History

4 Commits
Nov 18, 2024
Nov 18, 2024
Nov 18, 2024
Nov 18, 2024
Nov 18, 2024
Nov 18, 2024
Nov 18, 2024
Nov 18, 2024
Nov 18, 2024
Nov 18, 2024
Nov 18, 2024
Dec 23, 2024

Repository files navigation

LaPose: Laplacian Mixture Shape Modeling for RGB-Based Category-Level Object Pose Estimation

Environment Setup

To install the required dependencies, use the following commands:

conda env create -f Lapose.yaml

Data Preparation

  • Download the data from NOCS.
  • Download the segmentation predictions on CAMERA25 and REAL275 from DualPose-Net

Run the following scripts to prepare training instances:

python prepare_data/pose_data.py
python prepare_data/shape_data.py

Change the "dataset_dir" in config/config.py to your dataset directory.

Train

  • Train on the CAMERA+Real dataset.
python engine/train.py --model_save="./output/model_save"
  • Train on the CAMERA dataset.
python engine/train.py  --model_save="./output/model_save_CAMERA" --dataset=CAMERA
  • Train scale net.
python engine/train_scale_net.py --model_save="./output_scale_net/model_save"

Evaluate

  • Evaluate on the Real dataset. We provide our model checkpoint for evaluation here.
python evaluation/evaluate.py --resume_model="./output/model_save/model.pth" --dataset=Real --use_scale_net --sn_path='./output_scale_net/model_save/model.pth'
  • Evaluate on the CAMERA dataset.
python evaluation/evaluate.py --resume_model="./output/model_save_CAMERA/model.pth" --dataset=CAMERA --use_scale_net --sn_path='./output_scale_net/model_save/model.pth'

Citation

If you find our work useful, please cite:

@inproceedings{zhang2024lapose,
  title={LaPose: Laplacian Mixture Shape Modeling for RGB-Based Category-Level Object Pose Estimation},
  author={Zhang, Ruida and Huang, Ziqin and Wang, Gu and Zhang, Chenyangguang and Di, Yan and Zuo, Xingxing and Tang, Jiwen and Ji, Xiangyang},
  booktitle={European Conference on Computer Vision},
  pages={467--484},
  year={2024},
  organization={Springer}
}

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published