Min-Cut/Max-Flow Problem Instances for Benchmarking This is a collection of min-cut/max-flow problem instances that can be used for benchmarking min-cut/max-flow algorithms. The collection is released in companionship with the paper: Jensen et al., “Review of Serial and Parallel Min-Cut/Max-Flow Algorithms for Computer Vision”. The problem instances are collected from a wide selection of sources to be as representative as possible. Specifically, this collection contains: Many of the problem instances (some are unavailable due to dead links) published by the University of Waterloo at https://vision.cs.uwaterloo.ca/data/maxflow : Stereo problems based on [B98] and [K01]. 3D Segmentation problems based on [B01], [B06a], [B03]. Multi-view reconstruction problems based on [L06] and [B06b]. Surface fitting problems based on [L07]. Problem instances from from Verma’s & Batra’s review paper [V12]: Super resolution based on [F00] and [R07]. Texture restoration based on [R07]. Deconvolution based on [R07]. Decision tree field (DTF) based on [N11]. Automatic labelling environment (ALE) based on [E10], [ALE], [L09], and [L10]. Sparse Layered Graph (SLG) problems from [J20a]. Multi object surface fitting problems from [J20b]. Deep LOGISMOS surface fitting problem based on [G18]. Oriented MRF segmentation based on [B04], [R21], [E14]. U-Net segmentation cleaning with MRFs based on [B04] “An Experimental Comparison of Min-Cut/Max-Flow Algorithms for Energy Minimization in Vision”, 2004, PAMI: Cleaning of V-Net segmentations based on [R21], [M16], [E14]. Cleaning of U-Net segmentations based on [S19], [C16]. Mesh segmentation problems based on [L15]. Graph matching problems from [H21]. The orignal matching problems can be found at https://vislearn.github.io/libmpopt/iccv2021/. Here, we publish the QPBO subproblems for each matching problem to be used for benchmarking: Wide baseline matching based on [T08] and [C09]. Key point matching based on [E10] and [L12]. Large displacement flow based on [A15], [S17]. OpenGM matching based on [K08], [K15]. Worm atlas matching based on [K14]. Worm-to-worm matching based on [H12]. The reason for releasing this collection is to provide a single place download all datasets used in our paper (and various previous paper) instead of having to scavenge from multiple sources. Furthermore, several of the problem instances typically used for benchmarking min-cut/max-flow algorithms are no longer available at their original locations and may be difficult to find. By storing the data with a dedicated DOI we hope to avoid this. For license information, please see the README. Files and formats We provide all problem instances in two file formats: DIMACS and a costum binary format. Each file has been zipped, and similar files have then been grouped into their own zip file (i.e., it is a zip of zips). DIMACS files have been prefixed with dimacs_ and binary files have been prefixed with bin_. For additional information on the file formats, please the see the README file.
References Please see the README file.

Min-Cut/Max-Flow Problem Instances for Benchmarking This is a collection of min-cut/max-flow problem instances that can be used for benchmarking min-cut/max-flow algorithms. The collection is released in companionship with the paper: Jensen et al., “Review of Serial and Parallel Min-Cut/Max-Flow Algorithms for Computer Vision”. The problem instances are collected from a wide selection of sources to be as representative as possible. Specifically, this collection contains: Many of the problem instances (some are unavailable due to dead links) published by the University of Waterloo at https://vision.cs.uwaterloo.ca/data/maxflow : Stereo problems based on [B98] and [K01]. 3D Segmentation problems based on [B01], [B06a], [B03]. Multi-view reconstruction problems based on [L06] and [B06b]. Surface fitting problems based on [L07]. Problem instances from from Verma’s & Batra’s review paper [V12]: Super resolution based on [F00] and [R07]. Texture restoration based on [R07]. Deconvolution based on [R07]. Decision tree field (DTF) based on [N11]. Automatic labelling environment (ALE) based on [E10], [ALE], [L09], and [L10]. Sparse Layered Graph (SLG) problems from [J20a]. Multi object surface fitting problems from [J20b]. Deep LOGISMOS surface fitting problem based on [G18]. Oriented MRF segmentation based on [B04], [R21], [E14]. U-Net segmentation cleaning with MRFs based on [B04] “An Experimental Comparison of Min-Cut/Max-Flow Algorithms for Energy Minimization in Vision”, 2004, PAMI: Cleaning of V-Net segmentations based on [R21], [M16], [E14]. Cleaning of U-Net segmentations based on [S19], [C16]. Mesh segmentation problems based on [L15]. Graph matching problems from [H21]. The orignal matching problems can be found at https://vislearn.github.io/libmpopt/iccv2021/. Here, we publish the QPBO subproblems for each matching problem to be used for benchmarking: Wide baseline matching based on [T08] and [C09]. Key point matching based on [E10] and [L12]. Large displacement flow based on [A15], [S17]. OpenGM matching based on [K08], [K15]. Worm atlas matching based on [K14]. Worm-to-worm matching based on [H12]. The reason for releasing this collection is to provide a single place download all datasets used in our paper (and various previous paper) instead of having to scavenge from multiple sources. Furthermore, several of the problem instances typically used for benchmarking min-cut/max-flow algorithms are no longer available at their original locations and may be difficult to find. By storing the data with a dedicated DOI we hope to avoid this. For license information, please see the README. Files and formats We provide all problem instances in two file formats: DIMACS and a costum binary format. Each file has been zipped, and similar files have then been grouped into their own zip file (i.e., it is a zip of zips). DIMACS files have been prefixed with dimacs_ and binary files have been prefixed with bin_. For additional information on the file formats, please the see the README file.
References Please see the README file.

This dataset contains data, notebooks and code used in the publication: [1] Jeppesen, N., V.A. Dahl, A.N. Christensen, A.B. Dahl, L.P. Mikkelsen, Characterization of the fiber orientations in non-crimp glass fiber reinforced composites using structure tensor. IOP Conf. Ser.: Mater. Sci. Eng. 942, 012037, https://doi.org/10.1088/1757-899X/942/1/012037, 2020 If you reference this dataset, please also consider referencing the paper above. HF401TT-13_FoV16.5_Stitch.zip contains an X-ray CT scan of a non-crimp glass fiber composite sample saved in the TXM file format. HF401TT-13_FoV16.5_Stitch.txm.nii contains a cut-out of the TXM data, where air around the sample has been removed and the result has been saved in the NIfTI file format. The two notebooks, StructureTensorFiberAnalysisDemo and StructureTensorFiberAnalysisAdvancedDemo rely on the NIfTI scan data, HF401TT-13_FoV16.5_Stitch.txm.nii. They demonstrate how to do structure tensor orientation analysis on the data. The HF401TT-13_FoV16.5_Stitch notebook use the TXM scan data, HF401TT-13_FoV16.5_Stitch.zip. It can be used to recreate the results of the published experimental results. To run the notebooks, the Python file, structure_tensor_workers.py, must be in the same directory as the notebook. By default, the notebooks expect the following folder structure: /notebooks: Folder with the notebooks and Python files.
/originals: Folder with the data (TXM/NIfTI files).
/tmp: Folder for temporary files and output generated running the notebooks.
/notebooks/figures: Folder for exporting figures as files (only needed if you want to save figures as files).

This dataset contains data, notebooks and code used in the publication: [1] Jeppesen, N., V.A. Dahl, A.N. Christensen, A.B. Dahl, L.P. Mikkelsen, Characterization of the fiber orientations in non-crimp glass fiber reinforced composites using structure tensor. IOP Conf. Ser.: Mater. Sci. Eng. 942, 012037, https://doi.org/10.1088/1757-899X/942/1/012037, 2020 If you reference this dataset, please also consider referencing the paper above. HF401TT-13_FoV16.5_Stitch.zip contains an X-ray CT scan of a non-crimp glass fiber composite sample saved in the TXM file format. HF401TT-13_FoV16.5_Stitch.txm.nii contains a cut-out of the TXM data, where air around the sample has been removed and the result has been saved in the NIfTI file format. The two notebooks, StructureTensorFiberAnalysisDemo and StructureTensorFiberAnalysisAdvancedDemo rely on the NIfTI scan data, HF401TT-13_FoV16.5_Stitch.txm.nii. They demonstrate how to do structure tensor orientation analysis on the data. The HF401TT-13_FoV16.5_Stitch notebook use the TXM scan data, HF401TT-13_FoV16.5_Stitch.zip. It can be used to recreate the results of the published experimental results. To run the notebooks, the Python file, structure_tensor_workers.py, must be in the same directory as the notebook. By default, the notebooks expect the following folder structure: /notebooks: Folder with the notebooks and Python files.
/originals: Folder with the data (TXM/NIfTI files).
/tmp: Folder for temporary files and output generated running the notebooks.
/notebooks/figures: Folder for exporting figures as files (only needed if you want to save figures as files).

Data for CVPR 2020 paper Sparse Layered Graphs for Multi-Object Segmentation.

Related notebooks can be found here:
http://doi.org/10.11583/DTU.12016941
data.zipContains images used in the notebooks and paper experiments.
NT32_tomo3_.raw
This is a micro-CT scan of hand nerves used in the paper. The volume is 2048x2048x2048 uint16.
NT32_cLineLabel_scale4_preSegm_v3.nii.gz
This is a segmentation of the center lines of 216 nerves in the NT32 scan. The segmentation is made at four times lower resolution than NT32_tomo3_.raw and has a size of 512x512x512.
The NT32 dataset was previously used in the paper Three-dimensional architecture of human diabetic peripheral nerves revealed by X-ray phase contrast holographic nanotomography.

Data for CVPR 2020 paper Sparse Layered Graphs for Multi-Object Segmentation.

Related notebooks can be found here:
http://doi.org/10.11583/DTU.12016941
data.zipContains images used in the notebooks and paper experiments.
NT32_tomo3_.raw
This is a micro-CT scan of hand nerves used in the paper. The volume is 2048x2048x2048 uint16.
NT32_cLineLabel_scale4_preSegm_v3.nii.gz
This is a segmentation of the center lines of 216 nerves in the NT32 scan. The segmentation is made at four times lower resolution than NT32_tomo3_.raw and has a size of 512x512x512.
The NT32 dataset was previously used in the paper Three-dimensional architecture of human diabetic peripheral nerves revealed by X-ray phase contrast holographic nanotomography.

Data for CVPR 2020 paper Sparse Layered Graphs for Multi-Object Segmentation.

Related notebooks can be found here:
http://doi.org/10.11583/DTU.12016941
data.zipContains images used in the notebooks and paper experiments.
The complete 3D volume used in the paper cannot be made publicly available at this time.