LorenaLiu6/deepstereo-checkpoint-sceneflow

DeepStereo Checkpoint (SceneFlow)

Pretrained checkpoint for the project:

Learned binocular-encoding optics for RGBD imaging using joint stereo and focus cues

Official PyTorch implementation: https://github.com/Lorena-Y-Liu/Learned-Binocular-Optics

Links

• Source code: https://github.com/Lorena-Y-Liu/Learned-Binocular-Optics
• Checkpoint repository: https://huggingface.co/LorenaLiu6/deepstereo-checkpoint-sceneflow

Files

• ds-model.ckpt: pretrained checkpoint trained on SceneFlow

Usage

First clone the source code repository:

git clone https://github.com/Lorena-Y-Liu/Learned-Binocular-Optics.git
cd Learned-Binocular-Optics

Then download the checkpoint from this Hugging Face repository and place it in a local checkpoint directory, for example:

mkdir -p checkpoints

Put the file at:

checkpoints/ds-model.ckpt

Then load the checkpoint according to your training or inference pipeline in the source code repository.

Citation

If you find this work useful, please cite the corresponding paper:

@inproceedings{ou2025learned,
  title={Learned binocular-encoding optics for RGBD imaging using joint stereo and focus cues},
  author={Ou, Liangxun and others},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year={2025}
}

Notes

This repository contains model weights only.

For training, evaluation, and implementation details, please refer to the official GitHub source code repository: https://github.com/Lorena-Y-Liu/Learned-Binocular-Optics