Software
The page lists AI & DAS software products developed by the MIDAS lab and our collaborations with other research groups.
01
Unsupervised regression learning using sparse representation
[2] David Hong, Caroline Crockett, and Il Yong Chun,
"Convolutional operator learning (for Julia),"
[GitHub] https://github.com/dahong67/ConvolutionalOperatorLearning.jl, 2019.
A Julia toolbox for multi-dimensional convolutional analysis operator learning (CAOL) potentially with hand-crafted filters.
[1] Il Yong Chun,
“CONVOLT: CONVolutional Operator Learning Toolbox (for Matlab),”
[GitHub] https://github.com/mechatoz/convolt, 2019.
CONVOLT includes codes for fast and convergent convolutional analysis operator learning (CAOL) and convolutional dictionary learning (CDL). The codes are developed mainly for learning 2D convolutional regularizers from large datasets.
02
Iterative neural networks
We are preparing GitHub repositories.
03
Light-field imaging using focal stack
[1] Miao-Bin Lien,
"Light-field photography using focal stacks,"
[GitHub] https://github.com/miaobinlien/ranging-and-light-field-imaging-with-transparent-photodetectors, 2019.
A Matlab toolbox for light-filed photography using focal stacks. The codes were applied to light-field imaging using transparent photodetectors; see our Nature Photonics 14(3):143–148 (2020) paper.