NOTE

We hope that by sharing our data and software with our colleagues we start a culture of collaborative research.

All material presented here is either protected by patents or copyright. The data and software here are only for research purposes (For licencing please contact Duke University's Office of Licensing & Ventures (OLV)).

Please cite our corresponding paper if you use this material in any form in your publication. You may not redistribute our material without our written permission.

Vision and Image Processing (VIP) Laboratory

 

DISCLAIMER: “All images included in this website have been fully de-identified. Any dates associated with the imaging files do not relate to the subject or to date of image acquisition. Images are intended for use in research and educational settings.   Commercialization/redistribution of the images is prohibited.    In the unlikely event that you identify any remaining identifiers in the images, you are prohibited from further disclosure and should destroy all copies of the image and immediately notify the owner of this webpage at: sina.farsiu@duke.edu   All use of the images should include citation and credit to their corresponding paper.”

Data Sets:

Through the years, according to the reproducible research principal, me and my colleagues have made available almost all data sets used in our publications. The condition for using any of these data-sets in your research is referencing our corresponding publication. 


1:
World's largest annotated SD-OCT dataset for intermediate AMD and control subjects used in our recent paper  
Individual SDOCT images and marking: 35400 BScans from 269 AMD patients and 115 normal subjects, their ages, and their corresponding segmentation boundaries on a 5mm diameter centered at the fovea are available here.

2: Dataset for Denoising and interpolation Ophthalmic SDOCT Images (human and mice),
Raw data as well as dataset with results from our sparsity based novel sparsity based simultaneous denoising and interpolation (SBSDI) method (TMI, 2013) as compared to other state of the art denoising algorithms are available here.

3: Dataset for Denoising Ophthalmic SDOCT Images
Raw data as well as dataset with results from our sparsity based novel multiscale sparsity based tomographic denoising (MSBTD) method (BOE, 2012) as compared to other state of the art denoising algorithms are available here

4: Dataset for Classification of Ophthalmic SD-OCT Images of Normal, diabetic macular edema, and dry age-related macular degeneration subjects
This dataset consists of volumetric scans acquired from 45 patients: 15 normal patients, 15 patients with dry AMD, and 15 patients with DME using Spectralis SD-OCT (Heidelberg Engineering Inc.) are available here

5: Dataset for Superresolution and demosaicing

Most data sets that I have created for different super-resolution and demosaicing scenarios are available here. 

6: Dataset for SD-OCT Retinal Image Analysis  

Dataset, with automatic and manual annotations used in our recent automated retinal image analysis paper are available here.

7: Dataset for Kernel Regression Kernel Regression Image Processing

8: Dataset for Confocal Microscopy RPE Cell Analysis  

Dataset, with automatic and manual annotations used in our recent automated closed contour feature segmentation paper are available here.

9: Dataset for AO-SLO cone photoreceptor automatic segmentation and Analysis  

Dataset, with automatic and manual annotations used in our recent automated AOSLO cone segmentation paper are available here.

10: Dataset for FA leakage segmentation and Analysis in DME eyes  

Dataset, with manual annotations used in our recent automated segmentation of fluorescein leakage in subjects with diabetic macular edema paper are available here.

11: Dataset for OCT cyst and layer segmentation and Analysis in DME eyes  

Dataset, with manual annotations used in our recent automated segmentation of OCT images in subjects with diabetic macular edema paper are available here.

12: Dataset for Artery/Vein Classification   

Dataset, with manual annotations used in our recent paper on artery/vein classification on wide field-of-view (as well as classic small field of view) retinal images are available here.

13: Dataset for Tree Topology Estimation   

Dataset, with manual annotations used in our recent paper on estimation of 3-D topology from a single 2-D image of wide firld-of view color images of retina, images of rice roots, and synthetic trees are available here.

14: Dataset for Compressed Wavefront Sensing  

Dataset used in our recent paper on fast and efficient wavefront sensing are available here.

15: Open source software for automatic detection of cone photoreceptors in adaptive optics ophthalmoscopy using convolutional neural networks: Annotated dataset for detecting cones on both confocal and AOSLO images are available on GitHub (click on the link). This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution. Copyright © 2017 by Duke University

 

Software:

1.      Open source software for automatic detection of cone photoreceptors in adaptive optics ophthalmoscopy using convolutional neural networks: Source Code for our recent paper on detecting cones on both confocal and AOSLO images are available on GitHub (click on the link). This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution. Copyright © 2017 by Duke University

2. Open source Zemax optical design, Solidworks mechanical design, and LabView software, controlling the hardware of our Handheld Adaptive Optics Scanning Laser Ophthalmoscope.

3.      Automated Vessel Segmentation on Video Indirect Ophthalmology Software: Source Code for our recent paper are available here. This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution. Copyright © 2012 by Duke University

4.    Sparsity based simultaneous denoising and interpolation (SBSDI) :Software and dataset for our recent paper are available here. This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution. Copyright © 2013 by Duke University

5.     Sparse Zernike Representation (SPARZER) for Wavefront Sensing: Software and dataset for our recent paper are available here. This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution. Copyright © 2014 by Duke University

6.     Multiscale Sparsity Based Tomographic Denoising (MSBTD): Software and dataset for our recent paper are available here. This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution. Copyright © 2012 by Duke University

7.      MDSP Resolution Enhancement (Super Resolution) Software: According to the reproducible research principal, I created this (now commercialized) software package in collaboration with Drs. Dirk Robinson, Prof. Michael Elad, and Prof. Peyman Milanfar while at UCSC. The main function of this software tool is the implementation of several superresolution techniques in our published work, and the comparison of these to other work. The techniques implemented cover robust methods, dynamic color super-resolution methods, and simultaneous demosaicing and resolution enhancement. Please forward all your questions and enquiries directly to my former PhD advisor Prof. Milanfar, who is managing/distributing the software. This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution. Copyright © 2004 by University of California.

Also, due to its popularity, many other teams have also implemented our algorithms and have distributed open access code online. Here is an example implementation of our core method (Farsiu et al. IEEE TIP 2004) by an external investigator.

8.     Adjusted Mean Arc Length (AMAL) Graph Search source code: Get the source code from here. This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution. Copyright © 2016 by Duke University

9.     Kernel Regression code: Get the Kernel Regression Image Processing Toolbox and dataset from here. This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution. Copyright © 2007 by University of California

10.     Segmentation based sparse reconstruction (SSR) :Software for our recent paper are available here. This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution.

11.     Segmentation guided registration of wide field-of-view retinal optical coherence tomography volumes : Source code for our recent paper are available here. This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution.

12.     Software for color based segmentation of mullar glia cells : Source code for our recent paper are available here. This is experimental software. It is provided for non-commercial research purposes only. Use at your own risk. No warranty is implied by this distribution.

13. Open source software for our paper ""Statistical Models of Signal and Noise and Fundamental Limits of Segmentation Accuracy in Retinal Optical Coherence Tomography, IEEE TMI, 2018"

14. Open source software for our paper "Open-source, machine and deep learning-based automated algorithm for gestational age estimation through smartphone lens imaging"