This code is based on the paper: "Learned reconstructions for practical mask-based lensless imaging" available here: (https://www.osapublishing.org/oe/fulltext.cfm?uri=oe-27-20-28075&id=420747)
Clone this project using:
git clone https://github.com/Waller-Lab/LenslessLearning.git
The dependencies can be installed by using:
conda env create -f environment.yml
source activate lensless_learning
In addition, the LPIPS package is needed (this is used in the loss function during training). Instructions for installing LPIPS can be found here: (https://github.com/richzhang/PerceptualSimilarity)
The pre-trained models can be downloaded here
Jupyter Notebook: pre-trained reconstructions.ipynb
- Loads in the pre-trained models and runs reconstructions on sample lensless images.
- Initializes un-trained models and shows output images before training. Changes model parameters to the pre-loaded parameters and shows sample reconstructions
- The full Lensless Learning Dataset can be found here: (https://waller-lab.github.io/LenslessLearning/dataset.html)
- In addition, the 'in the wild' images taken without a computer monitor can be found here: (https://drive.google.com/drive/folders/1dtyxApqryiXbpqLSSUKreCVKfjQcT7pS?usp=sharing)