Skip to main content
ConductScienceScienceDex
Powered by ShareScore

Find research datasets worth reusing

Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.

56

datasets available to search

ShareScore release 0.7.1

Reset

Dataset results

56 results for “image recognition”

Learn how ShareScore rates datasets ↗
zenodo44/100

1QIsaa data collection (binarized images, feature files, and plotting scripts) for writer identification test using artificial intelligence and image-based pattern recognition techniques

<p><strong>The Great Isaiah Scroll (1QIsa<sup>a</sup>) data set for writer identification</strong></p> <p>This data set is collected for the ERC project:<br> The Hands that Wrote the Bible: Digital Palaeography and Scribal Culture of the Dead Sea Scrolls<br> PI: Mladen Popović<br> Grant agreement ID: 640497</p> <p>Project website: <a href="https://cordis.europa.eu/project/id/640497">https://cordis.europa.eu/project/id/640497</a><br> <br> <strong>Copyright (c) </strong>&nbsp;&nbsp; &nbsp;University of Groningen, 2021. All rights reserved.<br> <strong>Disclaimer and copyright notice for all data contained on this .tar.gz file:</strong></p> <p><strong>1)</strong> permission is hereby granted to use the data for research purposes. It is not allowed to distribute this data for commercial purposes.</p> <p><strong>2) </strong>provider gives no express or implied warranty of any kind, and any implied warranties of merchantability and fitness for purpose are disclaimed.</p> <p><strong>3) </strong>provider shall not be liable for any direct, indirect, special, incidental, or consequential damages arising out of any use of this data.</p> <p><strong>4) </strong>the user should refer to the first public article on this data set:<br> <br> <em>Popović, M., Dhali, M. A., &amp; Schomaker, L. (2020). Artificial intelligence-based writer identification generates new evidence for the unknown scribes of the Dead Sea Scrolls exemplified by the Great Isaiah Scroll (1QIsa<sup>a</sup>). arXiv preprint arXiv:2010.14476.</em><br> <br> BibTeX:</p> <pre>@article{popovic2020artificial, title={Artificial intelligence based writer identification generates new evidence for the unknown scribes of the Dead Sea Scrolls exemplified by the Great Isaiah Scroll (1QIsaa)}, author={Popovi{\&#39;c}, Mladen and Dhali, Maruf A and Schomaker, Lambert}, journal={arXiv preprint arXiv:2010.14476}, year={2020} }</pre> <p><strong>5) </strong>the recipient should refrain from proliferating the data set to third parties external to his/her local research group. Please refer interested researchers to this site for obtaining their own copy.</p> <p><strong>Organisation of the data:</strong></p> <p>The .tar.gz file contains three directories: images, features, and plots. The included &#39;README&#39; file contains all the instructions.</p> <p>The &#39;images&#39; directory contains NetPBM images of the columns of 1QIsa<sup>a</sup>. The NetPBM format is chosen because of its simplicity. Additionally, there is no doubt about lossy compression in the processing chain. There are two images for each of the Great Isaiah Scroll columns: one is the direct binarized output from the BiNet (<em>arxiv.org/abs/1911.07930</em>) system, and the other one is the manually cleaned version of the binarized output. &nbsp; The file names for the direct binarized output are of the format &#39;1QIsaa_col&lt;columnnr&gt;.pbm&#39;, for example, &#39;1QIsaa_col15.pbm&#39;. And, for the cleaned version, the format is &#39;1QIsaa_col&lt;columnnr&gt;_cleaned.pbm&#39;, for example, &#39;1QIsaa_col15_cleaned.pbm&#39;. Note: the image files are not in a separate directory; they will be extracted in the same place. However, due to the unique naming, there is no problem extracting them in one single directory.</p> <p>The &#39;features&#39; directory contains feature files computed for each of the column images. There are two types of feature files: Hinge and Adjoined. They are distinguishable by their extension, for example, &#39;1QIsaa_col15_cleaned.hinge&#39; and &#39;1QIsaa_col15_cleaned.adjoined&#39;. They are also arranged in separate directories for ease of use.</p> <p>The &#39;plots&#39; directory contains a simple python script to perform PCA on the feature files and then visualize them in a 3D plot. The file takes the location of feature files as an input. The &#39;README_plot&#39; file contains examples of how-to-run in the terminal.</p> <p><strong>Brief description:</strong><br> According to ImageMagick&#39;s&#39; identify&#39; tool, the original images are in grayscale (.jpg) from Brill collection, in &#39;8-bit Gray 256c&#39;. &nbsp;These images pass through multiple preprocessing measures to become suitable for pattern recognition-based techniques. The first step in preprocessing is the image-binarization technique. In order to prevent any classification of the text-column images based on irrelevant background patterns, a specific binarization technique (BiNet) was applied, keeping the original ink traces intact. After performing the binarization, the images were cleaned further by removing the adjacent columns that partially appear on the target columns&#39; images. Finally, few minor affine transformations and stretching corrections were performed in a restrictive manner. These corrections are also targeted for aligning the texts where the text lines get twisted due to the leather writing surface&#39;s degradation. Hence, the clean images are there in the directory along with the direct binarized images. No effort has been made to obtain a balanced set in any way.</p> <p><strong>Tools:</strong><br> <strong>Binarization:</strong><br> The BiNet tool is available for scientific use upon request (m.a.dhal(at)rug.nl)</p> <p><strong>Image Morphing:</strong><br> In the original article, data augmentation was performed using image morphing. The tool is available on GitHub:<br> https://github.com/GrHound/imagemorph.c</p> <p><strong>Features for writer identification:</strong><br> Lambert Schomaker<br> http://www.ai.rug.nl/~lambert/allographic-fraglet-codebooks/allographic-fraglet-codebooks.html<br> http://www.ai.rug.nl/~lambert/hinge/hinge-transform.html<br> <em><strong>1.&nbsp;</strong>L. Schomaker &amp; M. Bulacu (2004). Automatic writer identification using connected-component contours and edge-based features of upper-case Western script. IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol 26(6), June 2004, pp. 787 - 798.<br> <strong>2. </strong>Bulacu, M. &amp; Schomaker, L.R.B. (2007). Text-independent Writer Identification and Verification Using Textural and Allographic Features, &nbsp;IEEE Trans. on Pattern Analysis and Machine Intelligence (PAMI), Special Issue - Biometrics: Progress and Directions, April, 29(4), p. 701-717.</em><br> &nbsp;<br> The features (hinge, fraglets) have been combined in a single MS Windows application, GIWIS, which is available for scientific use upon request (l.r.b.schomaker(at)rug.nl)</p> <p><strong>If you have any question, please contact us:</strong><br> Maruf A. Dhali &lt;m.a.dhali(at)rug.nl&gt;<br> Lambert Schomaker &lt;l.r.b.schomaker(at)rug.nl&gt;<br> Mladen Popović &lt;m.popovic(at)rug.nl&gt;</p> <p><strong>Please cite our papers if you use this data set:</strong><br> <em><strong>1.</strong> Popović, M., Dhali, M. A., &amp; Schomaker, L. (2020). Artificial intelligence based writer identification generates new evidence for the unknown scribes of the Dead Sea Scrolls exemplified by the Great Isaiah Scroll (1QIsa<sup>a</sup>). arXiv preprint arXiv:2010.14476.<br> <strong>2. </strong>Dhali, M. A., de Wit, J. W., &amp; Schomaker, L. (2019). Binet: Degraded-manuscript binarization in diverse document textures and layouts using deep encoder-decoder networks. arXiv preprint arXiv:1911.07930.</em></p>

opencc-by-4.0Jan 2021View details →
zenodo44/100

Firemaker image collection for benchmarking forensic writer identification using image-based pattern recognition

<p>Disclaimer and terms of use:<br> ============================</p> <p>/*****************************************************************************\<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; This is the Firemaker NFI-images Distribution &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; This distribution contains 1000 images of scanned handwritten text, &nbsp; &nbsp; &nbsp; *<br> * &nbsp; scanned at resolution 300dpi grey scale, containing pages of &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;*<br> * &nbsp; handwritten text by 250 writers, four pages per writer, from four &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; writing conditions, one condition per page. The conditions are: &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; p1: copied, natural style, p2: copied, UPPER case, p3: copied and forged, *<br> * &nbsp; i.e.,&quot;try to write in a different style than your natural style&quot;, and p4, *<br> * &nbsp; self generated, i.e., text produced to describe a given cartoon. &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;*<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; Copyright The International Unipen Foundation, 2000, All rights reserved &nbsp;*<br> *******************************************************************************<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp;DISCLAIMER AND COPYRIGHT NOTICE FOR ALL DATA CONTAINED ON THIS CDROM: &nbsp; &nbsp; &nbsp;*<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp;1) PERMISSION IS HEREBY GRANTED TO USE THE DATA FOR RESEARCH &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; PURPOSES. IT IS NOT ALLOWED TO DISTRIBUTE THIS DATA FOR COMMERCIAL &nbsp; &nbsp; &nbsp;*<br> * &nbsp; &nbsp; PURPOSES. &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp;2) PROVIDER GIVES NO EXPRESS OR IMPLIED WARRANTY OF ANY KIND AND ANY &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR PURPOSE ARE &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; DISCLAIMED. &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp;3) PROVIDER SHALL NOT BE LIABLE FOR ANY DIRECT, INDIRECT, SPECIAL, &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF ANY USE OF THIS &nbsp; &nbsp; &nbsp;*<br> * &nbsp; &nbsp; DATA. &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp;4) THE USER SHOULD REFER TO THE FIRST PUBLIC ARTICLE ON THIS DATA SET: &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; M. Bulacu, L. Schomaker &amp; L. Vuurpijl (2003). &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; Writer identification using edge-based directional features. &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;*<br> * &nbsp; &nbsp; ICDAR &#39;03: Proceedings of the 7th International Conference on Document &nbsp;*<br> * &nbsp; &nbsp; Analysis and Recognition, pp. 937-941. &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;*<br> * &nbsp; &nbsp; Piscataway: IEEE Computer, ISBN 0-7695-1960-1 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> * &nbsp;5) THE RECIPIENT SHOULD REFRAIN FROM PROLIFERATING THE DATA SET TO THIRD &nbsp; *<br> * &nbsp;PARTIES EXTERNAL TO HIS/HER LOCAL RESEARCH GROUP. PLEASE REFER INTERESTED &nbsp;*<br> * &nbsp;RESEARCHERS TO HTTP://UNIPEN.ORG FOR OBTAINING THEIR OWN COPY. &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; *<br> \*****************************************************************************/</p> <p>BibTeX entry: &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;&nbsp;</p> <p>&nbsp; @inproceedings{Firemaker, &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;<br> &nbsp; &nbsp; author = {Bulacu, M. and Schomaker, L.R.B. and Vuurpijl, L.}, &nbsp; &nbsp;<br> &nbsp; &nbsp; title = {Writer Identification Using Edge-Based Directional Features},<br> &nbsp; &nbsp; booktitle = {ICDAR &#39;03: Proceedings of the 7th International&nbsp;<br> &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;Conference on Document Analysis and Recognition},<br> &nbsp; &nbsp; year = {2003},<br> &nbsp; &nbsp; isbn = {0-7695-1960-1},<br> &nbsp; &nbsp; pages = {937-941},<br> &nbsp; &nbsp; publisher = {IEEE Computer Society},<br> &nbsp; &nbsp; address = {Washington, DC, USA},<br> &nbsp; &nbsp;}</p> <p>In the project &quot;Vergelijk&quot;, a grant obtained from the Dutch Forensic Science<br> Institute, two existing professional writer-identification systems have been&nbsp;<br> compared regarding usability studies and in particular recognition&nbsp;<br> performance (Schomaker &amp; Vuurpijl, 2000). The results of this comparison&nbsp;<br> are contained in a confidential report:</p> <p>&nbsp; L.R.B. Schomaker and L.G. Vuurpijl (2000).&nbsp;<br> &nbsp; Forensic writer identification: A benchmark data set&nbsp;<br> &nbsp; and a comparison of two systems. Technical report,&nbsp;<br> &nbsp; Nijmegen Institute for Cognition and Information (NICI),&nbsp;<br> &nbsp; University of Nijmegen, The Netherlands.</p> <p>Informative and non-confidential details from this report are&nbsp;<br> given in the accompanying file: &nbsp;&#39;firemaker-dbase.pdf&#39;</p> <p>To compare both systems, a carefully designed experiment was conducted to<br> record handwritten samples from male and female writers in several conditions:</p> <p>Condition 1: Normal constrained handwriting<br> ==============================================</p> <p>Below, the Dutch text writers had to produce in normal handwriting is given.&nbsp;</p> <p>--- start text ----<br> Zij bezochten veilingen en reisden met de KLM. Voor<br> korte afstanden huurden ze een auto, meestal een VW<br> of een Ford.<br> &lt;EMPTY LINE&gt;<br> De veilingen waren van 7-4-1993 tot 3-5-1993 in New<br> York, Tokyo, Qu&eacute;bec, Rome, Parijs, Z&uuml;rich en Oslo.<br> &lt;EMPTY LINE&gt;<br> Omdat de veilingen steeds begonnen om 12 uur en je<br> gemiddeld 200 tot 300 kilometer moest rijden,<br> stonden zij steeds om 6.30 uur op en vertrokken om<br> 8 uur uit het hotel.<br> &lt;EMPTY LINE&gt;<br> Elke dag hadden ze vijfhonderd (f 500,-) gulden<br> nodig. Daarvoor gebruikten ze elke keer een cheque<br> van tweehonderd (f 200,-) en een cheque van<br> driehonderd (f 300,-) gulden. Aan geschenken gaven<br> ze ongeveer honderd gulden (f 100,-) uit.<br> --- end text ----</p> <p><br> Condition 2: Production of constrained block capital handwriting<br> ================================================================</p> <p>In this condition, the writers had to produce the following text<br> in block-capital handwriting:</p> <p>--- start text ----<br> NADAT ZE IN NEW YORK, TOKYO, QU&Eacute;BEC, PARIJS, Z&Uuml;RICH<br> EN OSLO WAREN GEWEEST, VLOGEN ZE UIT DE USA TERUG<br> MET VLUCHT KL 658 OM 12 UUR.<br> &lt;empty line&gt;<br> ZE KWAMEN AAN IN DUBLIN OM 7 UUR EN IN AMSTERDAM OM<br> 9.40 UUR &#39;S AVONDS. DE FIAT VAN BOB EN DE VW VAN<br> DAVID STONDEN IN R3 VAN HET PARKEERTERREIN.<br> HIERVOOR MOESTEN ZE HONDERD GULDEN (F 100,-)<br> BETALEN.<br> --- end text ----</p> <p><br> Condition 3: Production of free-forged handwriting<br> ==================================================</p> <p>Below, the text writers had to produce in the free-forged handwriting<br> condition is given. No example of handwriting is given which they have to<br> mimick (forge), the condition concerns a self-conceived distorted&nbsp;<br> handwriting style.</p> <p>--- start text ----<br> Nog dezelfde avond reden ze naar hun vrienden<br> Chris, Emile, Jan, Irene en Henk, nadat ze hun<br> vriendinnen Greta en Maria hadden opgehaald.<br> &lt;EMPTY LINE&gt;<br> Samen hadden ze vijfhonderd (500) zeldzame<br> postzegels gekocht, Bob driehonderd (300) en David<br> tweehonderd (200).<br> &lt;EMPTY LINE&gt;<br> De reis was de moeite waard geweest.<br> --- end text ----</p> <p><br> Condition 4: Production of unconstrained handwriting<br> ====================================================</p> <p>The final text writers had to produce is unconstrained handwriting.<br> The cartoon, a series of pictures concerning a &#39;UFO&#39; landing had<br> to be described in their own words, in at least six lines of text.<br> See image file &quot;space.gif&quot;.</p> <p><br> Thruth labels and writer identifications<br> ========================================</p> <p>Each writer has a unique id, specified as:</p> <p>&nbsp; &nbsp;id: &nbsp; {num}{set}<br> &nbsp; num: &nbsp; a three-digit number<br> &nbsp;set: &nbsp; &nbsp;either 01, 02, 03 or 04, identifying one of the 4 experiments</p> <p>The vast majority of the writers producing sets 01, 02 and 03 mimicked the<br> content and layout (empty lines) of the constrained texts they had to copy<br> sufficiently accurately, such that the example texts are a good indication of<br> the contents. However, as set 04 (&quot;describe cartoon story&quot;) &nbsp;contains<br> unconstrained self-generated handwriting, the corresponding thruth &nbsp;labels had<br> to be extracted manually. The resulting label files are contained in &nbsp;the<br> directory ./300dpi/p4-self-natural/labels/</p> <p>Note: no letter, word, line or paragraph segmentation is provided with this<br> data set. The main text can be cropped easily. Since the orientation is<br> horizontal, projection techniques can be used to extract lines, using<br> a line-spacing parameter (~94 pixels line height) as an additional check.&nbsp;</p> <p><br> Overview of directories:</p> <p>300dpi/<br> &nbsp; &nbsp;p1-copy-normal/ &nbsp; &nbsp; &nbsp;Copying task, normal writing style &nbsp;<br> &nbsp; &nbsp;p2-copy-upper/ &nbsp; &nbsp; &nbsp; Copying task, UPPER-case&nbsp;<br> &nbsp; &nbsp;p3-copy-forged/ &nbsp; &nbsp; &nbsp;Copying task, instructed to mimic another script style<br> &nbsp; &nbsp;p4-self-natural/ &nbsp; &nbsp; Self-generated text, natural writing condition</p> <p>Note: the original raw collection contained writer #155, who has been removed<br> from this data set, as his first condition (p1) was started in upper case and<br> the page was not &nbsp;completed. Deleted files were 15501.tif, 15502.tif, 15503.tif<br> and 15504.tif.</p> <p>Note: the name of this data set (Firemaker) is a contraction of the names<br> Vuurpijl and Schomaker.</p> <p>Note b: Example of a cutout of essential handwritten text using NetPBM tools: &nbsp;<br> &nbsp;tifftopnm 15201.tif | pnmcut -left 50 -right 2400 -top 700 -bottom 3250 &gt; handwriting.pgm</p> <p>&nbsp;For an experiment, the upper and lower halves of the resulting image were<br> &nbsp;usually used in the Schomaker &amp; Bulacu studies to obtain two samples of&nbsp;<br> &nbsp;handwriting for a writer.</p> <p>&nbsp;http://www.ai.rug.nl/~lambert<br> &nbsp;http://www.ai.rug.nl/~bulacu</p> <p>Our features for writer identification:</p> <p>Lambert Schomaker<br> &nbsp;http://www.ai.rug.nl/~lambert/allographic-fraglet-codebooks/allographic-fraglet-codebooks.html<br> &nbsp;L. Schomaker &amp; M. Bulacu (2004).&nbsp;<br> &nbsp;Automatic writer identification using connected-component contours and edge-based features of upper-case Western script.&nbsp;<br> &nbsp;IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol 26(6), June 2004, pp. 787 - 798.</p> <p>Marius Bulacu<br> &nbsp;http://www.ai.rug.nl/~lambert/hinge/hinge-transform.html<br> &nbsp;Bulacu, M. &amp; Schomaker, L.R.B. (2007).&nbsp;<br> &nbsp;Text-independent Writer Identification and Verification Using Textural and Allographic Features,&nbsp;<br> &nbsp;IEEE Trans. on Pattern Analysis and Machine Intelligence (PAMI), Special Issue - Biometrics: Progress and Directions, April, 29(4), p. 701-717.</p> <p>Axel Brink<br> &nbsp;http://www.ai.rug.nl/~axel/ &nbsp;&#39;Quill&#39; feature<br> &nbsp;A.A. Brink, J. Smit, M.L. Bulacu, and L.R.B. Schomaker (2011).&nbsp;<br> &nbsp;Writer identification using directional ink-trace width measurements,&nbsp;<br> &nbsp;Pattern Recognition (July 2011), doi: 10.1016/j.patcog.2011.07.005<br> &nbsp;<br> These three feature groups (hinge, fraglets, quill) have been combined in<br> a single MS Windows application, GIWIS which is available for scientific<br> use upon request (schomaker@ai.rug.nl)</p> <p>Note c.</p> <p>The accompanying file &#39;Firemaker-writer-info.dat&#39; contains some<br> writer information:&nbsp;<br> Column 1: writer identification code<br> Column 2: sex<br> Column 3: handedness,&nbsp;<br> Column 4: age in years<br> Column 5: major Western script group (print,cursive or mixed)<br> &nbsp;</p>

opencc-by-4.0Dec 1999View details →
zenodo44/100

Raw images (photographs) of urban text scenes for camera-based Thai text recognition

<p>Raw image collection of city scenes in Thailand with text content.<br> Text is photographed from diffeent angles. Also morning and evening<br> photographs were taken in order to capture different lighting<br> conditions. The material, 309 images, was photographed in 2013 by<br> Bowornrat Sriman and volunteers.</p> <p>Example EXIF:<br> JPEG image data, Exif standard: [TIFF image data, little-endian,<br> direntries=13, height=2448, manufacturer=SAMSUNG, model=GT-I9300,<br> orientation=upper-right, xresolution=220, yresolution=228,<br> resolutionunit=2, software=I9300XXEMA2, datetime=2013:03:14 18:17:49,<br> GPS-Data, width=3264], baseline, precision 8, 3264x2448, frames 3</p> <p>The images are not labeled. The orientation (landscape/portrait) is<br> not corrected yet. This material was used in preparation of the publication:</p> <p>Sriman, B.&nbsp; &amp; Schomaker, L. (2015).<br> Object Attention Patches for Text Detection and Recognition in Scene Images using SIFT,<br> Proceedings of the International Conference on Pattern Recognition Applications and<br> Methods: ICPRAM 2015.&nbsp; De Marsico, M., Figueiredo, M.&nbsp; &amp; Fred, A.<br> (Eds.).&nbsp; Lisbon, Portugal: SciTePress, Vol.&nbsp; 1, p.&nbsp; 304-311 8 p.</p> <p>Please cite this publication when using these data.</p>

opencc-by-4.0Mar 2013View details →
zenodo40/100

CryoVirusDB: An Expert Labelled Cryo-EM Image Dataset for AI-Driven Virus Particle recognition and Extraction

<p><span>With the advancements in instrumentation, image processing algorithms, and computational capabilities, single-particle electron cryo-microscopy (cryo-EM) has achieved nearly atomic resolutions in the 3D reconstruction of viruses. These detailed structures play a crucial role in comprehending the biological functions and advancing the development of more precise vaccines and antiviral treatments. Despite the effectiveness of deep learning in analyzing microscopic images, its potential in identifying and extracting virus particles from cryo-EM micrographs has been hindered by the limited availability of diverse and high-quality datasets. In this study, we introduce 'CryoVirusDB,' a labeled dataset containing coordinates of accurately selected virus particles in cryo-EM micrographs. CryoVirusDB comprises 9,941 micrographs featuring 9 different viruses along with the coordinates of 0.2 million virus particles in total. We anticipate that CryoVirusDB will enhance the capabilities of deep learning in accurately identifying virus particles in cryo-EM micrographs, thereby facilitating the subsequent 2D-3D reconstruction process.</span></p> <p><span>Instructions to download and use dataset: https://github.com/BioinfoMachineLearning/CryoVirusDB</span></p>

opencc-by-4.0Dec 2023View details →
zenodo40/100

Dataset for the manuscript "Attribute Recognition: A New Method for Grouping Planetary Images by Visual Characteristics, Using the Example of Mn-Rich Rocks in the Floor of Gale Crater, Mars."

<p>This dataset supports the manuscript "Attribute Recognition: A New Method for Grouping Planetary Images by Visual Characteristics, Using the Example of Mn-Rich Rocks in the Floor of Gale Crater, Mars." The dataset is contained in a single CSV file with 201 data rows (one row per NASA Curiosity rover ChemCam instrument target used in the study). The columns in this dataset include the martian solar day (sol) on which each target was imaged by ChemCam; the standoff distance from ChemCam to each target (in meters); binary columns (values are either 1 or 0, indicating presence or absence, respectively) for each of the 17 visual attributes we documented for each target image; the corresponding greyscale ChemCam RMI mosaic file location (on the Planetary Data System); and columns indicating which group each target was sorted into under each classification algorithm discussed in the text (P_{SG}: simple graph method; P_{AP}: automatic partitioning method; P_{\lambda=1.6}: community detection method with \lambda=1.6). To obtain the binary strings used for the classification algorithms, the 17 visual attribute columns can be concatenated.&nbsp;</p> <p>Also included is a collection of HTML files that enables easy viewing of the RMI mosaics in each cluster, using the Planetary Data System links. To use it, download the <code>.zip</code> file, unzip it, and open the <code>index.html</code> file in the browser of your choice (likely will work to simply double-click <code>index.html</code>)</p>

opencc-by-4.0Jul 2024View details →
zenodo40/100

Animal Recognition Using Methods Of Fine-Grained Visual Analysis - Kashtanka Pets (All Dev and Test Images, Single Folder)

<p>Kashtanka Pets images, with all Dev and Test images (total 66639 images).&nbsp; In a single folder, with filenames indicating path of file in original dataset distribution.</p>

opencc-by-4.0Jul 2022View details →
zenodo40/100

The ORBIT (Object Recognition for Blind Image Training)-India Dataset

<div> <p>The ORBIT (Object Recognition for Blind Image Training) -India Dataset is a collection of 105,243 images of 76 commonly used objects, collected by 12 individuals in India who are blind or have low vision. This dataset is an "Indian subset" of the original ORBIT dataset [1, 2], which was collected in the UK and Canada. In contrast to the ORBIT dataset, which was created in a Global North, Western, and English-speaking context, the ORBIT-India dataset features images taken in a low-resource, non-English-speaking, Global South context, a home to 90% of the world&rsquo;s population of people with blindness. Since it is easier for blind or low-vision individuals to gather high-quality data by recording videos, this dataset, like the ORBIT dataset, contains images (each sized 224x224) derived from 587 videos. These videos were taken by our data collectors from various parts of India using the Find My Things [3] Android app. Each data collector was asked to record eight videos of at least 10 objects of their choice.&nbsp;</p> </div> <div> <p>Collected between July and November 2023, this dataset represents a set of objects commonly used by people who are blind or have low vision in India, including earphones, talking watches, toothbrushes, and typical Indian household items like a belan (rolling pin), and a steel glass. These videos were taken in various settings of the data collectors' homes and workspaces using the Find My Things Android app.&nbsp;</p> </div> <div> <p>The image dataset is stored in the &lsquo;Dataset&rsquo; folder, organized by folders assigned to each data collector (P1, P2, ...P12) who collected them. Each collector's folder includes sub-folders named with the object labels as provided by our data collectors. Within each object folder, there are two subfolders: &lsquo;clean&rsquo; for images taken on clean surfaces and &lsquo;clutter&rsquo; for images taken in cluttered environments where the objects are typically found. The annotations are saved inside a&nbsp; &lsquo;Annotations&rsquo; folder containing a JSON file per video (e.g., P1--coffee mug--clean--231220_084852_coffee mug_224.json) that contains keys corresponding to all frames/images in that video (e.g., "P1--coffee mug--clean--231220_084852_coffee mug_224--000001.jpeg": {"object_not_present_issue": false, "pii_present_issue": false}, "P1--coffee mug--clean--231220_084852_coffee mug_224--000002.jpeg": {"object_not_present_issue": false, "pii_present_issue": false}, ...). The &lsquo;object_not_present_issue&rsquo; key is True if the object is not present in the image, and the &lsquo;pii_present_issue&rsquo; key is True, if there is a personally identifiable information (PII) present in the image. Note, all PII present in the images has been blurred to protect the identity and privacy of our data collectors. This dataset version was created by cropping images originally sized at 1080 &times; 1920; therefore, an unscaled version of the dataset will follow soon.&nbsp;</p> </div> <div> <p>This project was funded by the Engineering and Physical Sciences Research Council (EPSRC) Industrial ICASE Award with Microsoft Research UK Ltd. as the Industrial Project Partner. We would like to acknowledge and express our gratitude to our data collectors for their efforts and time invested in carefully collecting videos to build this dataset for their community. The dataset is designed for developing few-shot learning algorithms, aiming to support researchers and developers in advancing object-recognition systems. We are excited to share this dataset and would love to hear from you if and how you use this dataset. Please feel free to reach out if you have any questions, comments or suggestions.&nbsp;</p> </div> <div> <p>REFERENCES:&nbsp;</p> </div> <div> <ol> <li> <p>Daniela Massiceti, Lida Theodorou, Luisa Zintgraf, Matthew Tobias Harris, Simone Stumpf, Cecily Morrison, Edward Cutrell, and Katja Hofmann. 2021. ORBIT: A real-world few-shot dataset for teachable object recognition collected from people who are blind or low vision. DOI: <a href="https://doi.org/10.25383/city.14294597" target="_blank" rel="noreferrer noopener">https://doi.org/10.25383/city.14294597</a></p> </li> <li> <p>microsoft/ORBIT-Dataset.&nbsp;<a href="https://github.com/microsoft/ORBIT-Dataset" target="_blank" rel="noreferrer noopener">https://github.com/microsoft/ORBIT-Dataset</a> &nbsp;</p> </li> <li> <p>Linda Yilin Wen, Cecily Morrison, Martin Grayson, Rita Faia Marques, Daniela Massiceti, Camilla Longden, and Edward Cutrell. 2024. Find My Things: Personalized Accessibility through Teachable AI for People who are Blind or Low Vision. In Extended Abstracts of the 2024 CHI Conference on Human Factors in Computing Systems (CHI EA '24). Association for Computing Machinery, New York, NY, USA, Article 403, 1&ndash;6.&nbsp;<a href="https://doi.org/10.1145/3613905.3648641" target="_blank" rel="noreferrer noopener">https://doi.org/10.1145/3613905.3648641</a>&nbsp;</p> </li> </ol> </div>

opencc-by-4.0Jun 2024View details →
zenodo40/100

BRAIN Journal-A Robust Approach of Facial Orientation Recognition from Facial Features-Figure 5. Sample Image data 2

<p>It helps to write our code in C# and to make an application in dot net framework, which collects facial images using a webcam/or other video grabbing tools. Then it implements Haar detection to extract facial features and to draw image pattern for matching both images.&nbsp;&nbsp;</p> <p>After image matching, we got a positive result at 93% times, for 1000 random sample images tested on the nine criteria of orientation.&nbsp;</p>

opencc-by-4.0Jul 2017View details →
zenodo40/100

BRAIN Journal-A Robust Approach of Facial Orientation Recognition from Facial Features-Figure 4. Sample image data 1

<p>After image matching, we got a positive result at 93% times, for 1000 random sample images tested on the nine criteria of orientation.&nbsp;</p> <p>After matching the images with the reference image, it gets the nearest orientation matches and they could be Font left, Font right, Down-left, Down Right, Up left, Upright, Font Straight, Up Straight, Down Straight. Initially, some constraints must be satisfied to realize a successful correct matching. The facial regions, concerned on eyes and nose points, have the following characteristic: if there is almost one missing point for the region of the same type then the comparison will be performed. There must be the same number of feature points for both eyes and nose separately. If this condition is satisfied then a new comparison will be performed.&nbsp;</p>

opencc-by-4.0Jul 2017View details →
zenodo40/100

BRAIN Journal-A Robust Approach of Facial Orientation Recognition from Facial Features-Figure 2. A Sample Image from extracted feature

<p>Using these data, it creates a new picture and uses these data as a starting point for drawing. By using the data, it gets a model and shape of face without color and facial expression (Gourier et al.; 2004), such as Figure 2. It got a model of faces using these features.&nbsp;</p>

opencc-by-4.0Jul 2017View details →
zenodo40/100

BRAIN Journal-A Robust Approach of Facial Orientation Recognition from Facial Features-Figure 5. Sample Image data 2

<p>After image matching, we got a positive result at 93% times, for 1000 random sample images tested on the nine criteria of orientation.&nbsp;</p> <p>Face orientation recognition is an important topic in computer vision and pattern recognition. Due to the non-rigid properties of faces, it is computationally expensive and difficult to achieve good recognition accuracy and robustness in face orientation recognition. In this paper, we propose an image mapping technique for face analysis in smart camera networks with a feature extraction and data from the facial feature. We estimate the face orientation angles in all camera views, based on the matched imaged data. Our objective is to obtain a set of facial structures which can work as landmarks for tracking and recognition of facial expressions.&nbsp;</p>

opencc-by-4.0Jul 2017View details →
zenodo40/100

BRAIN Journal-A Robust Approach of Facial Orientation Recognition from Facial Features-Figure 4. Sample image data 1

<p>After matching the images with the reference image, it gets the nearest orientation matches and they could be Font left, Font right, Down-left, Down Right, Up left, Upright, Font Straight, Up Straight, Down Straight. Initially, some constraints must be satisfied to realize a successful correct matching. The facial regions, concerned on eyes and nose points, have the following characteristic: if there is almost one missing point for the region of the same type then the comparison will be performed. There must be the same number of feature points for both eyes and nose separately. If this condition is satisfied then a new comparison will be performed.&nbsp;</p> <p>After image matching, we got a positive result at 93% times, for 1000 random sample images tested on the nine criteria of orientation.&nbsp;</p>

opencc-by-4.0Jul 2017View details →
zenodo40/100

BRAIN Journal-A Robust Approach of Facial Orientation Recognition from Facial Features-Figure 2. A Sample Image from extracted feature

<p>Using these data, it creates a new picture and uses these data as a starting point for drawing. By using the data, it gets a model and shape of face without color and facial expression (Gourier et al.; 2004), such as Figure 2. It got a model of faces using these features.</p>

opencc-by-4.0Jul 2017View details →
zenodo40/100

Performance analysis of micro-expression recognition over different sample image sizes.

<p>Performance of micro-expression recognition accuracy analyzed using different sample sizes with motion and geometric features. The sample sizes analyzed are: 140x170, 280x340, 560x680 and 1120x1360 using SMIC, CASMEII, CAS(ME)^2 and SAMM. The experiments were conducted using three different feature extraction setups: (i) optimized BiWOOF, (ii) full-face graph, and (iii) full-face graph with amplitude-based emotion magnification method (A-EMM). The results were compared with the results of the original BiWOOF presented in [1] as the baseline study. These files are published under CC0 license.</p> <p>&nbsp;</p> <p>References</p> <p>[1] Liong, S. T., See, J., Wong, K., &amp; Phan, R. C. W. (2018). Less is more: Micro-expression recognition from video using apex frame. <em>Signal Processing: Image Communication</em>, <em>62</em>, 82-92.</p>

opencc-zeroSep 2021View details →
zenodo40/100

Scaled and Translated Image Recognition (STIR)

<p><strong>Paper:</strong>&nbsp;<a href="https://arxiv.org/abs/2211.10288">[2211.10288] Just a Matter of Scale? Reevaluating Scale Equivariance in Convolutional Neural Networks (arxiv.org)</a><br> <strong>Code:</strong>&nbsp;<a href="https://github.com/taltstidl/scale-equivariant-cnn">taltstidl/scale-equivariant-cnn: Official code for &quot;Just a Matter of Scale? Reevaluating Scale Equivariance in Convolutional Neural Networks&quot; (github.com)</a></p> <p>While convolutions are known to be invariant to (discrete) translations, scaling continues to be a challenge and most image recognition networks are not invariant to them. To explore these effects, we have created the Scaled and Translated Image Recognition (STIR) dataset. This dataset contains objects of size <span class="math-tex">\(s \in [17,64]\)</span>, each randomly placed in a <span class="math-tex">\(64 \times 64\)</span>&nbsp;pixel image.</p> <p><strong>Using the dataset</strong></p> <p>Depending on which data you are planning to use, download one or more of the following files. Data is stored in compressed <code>.npz</code> format and can be loaded as documented <a href="https://numpy.org/doc/stable/reference/generated/numpy.load.html">here</a>.</p> <table> <thead> <tr> <th>File</th> <th>Description</th> </tr> </thead> <tbody> <tr> <td><code>emoji.npz</code></td> <td>Emoji vector icons rendered as white icon on black background</td> </tr> <tr> <td><code>mnist.npz</code></td> <td>Classic MNIST handwritten digits rescaled to varying sizes</td> </tr> <tr> <td><code>trafficsign.npz</code></td> <td>Traffic signs from street imagery downscaled to varying sizes</td> </tr> <tr> <td><code>aerial.npz</code></td> <td>Objects in aerial imagery downscaled to varying sizes</td> </tr> </tbody> </table> <p>Each file contains multiple arrays that can be accessed in a dictionary-like fashion. The keys are documented below, where <code>n</code> is the number of classes for a given file and <code>m</code> is the number of instances for each class. Both <code>emoji.npz</code> (36 classes, 1 instance) and <code>mnist.npz</code> (10 classes, 50 instances) are in black &amp; white while <code>trafficsign.npz</code> (16 classes, 25 instances) and <code>aerial.npz</code> (9 classes, 25 instances) are in color.</p> <table> <thead> <tr> <th>Key</th> <th>Shape</th> <th>Description</th> </tr> </thead> <tbody> <tr> <td><code>imgs</code></td> <td><code>(3, 48, n, m, 64, 64)</code> black &amp; white, <code>(3, 48, n, 64, 64, 3)</code> color</td> <td>Images grouped into 3 sets (training, validation, testing) and 48 different scales. Values will be in range <code>0</code> to <code>255</code>.</td> </tr> <tr> <td><code>lbls</code></td> <td><code>(3, 48, n, m)</code></td> <td>Indices referencing ground truth labels. See <code>lbldata</code> for descriptive names. Values will be in range <code>0</code> to <code>n - 1</code>.</td> </tr> <tr> <td><code>scls</code></td> <td><code>(3, 48, n, m)</code></td> <td>Known scales as given by bounding box size. Values will be in range <code>17</code> to <code>64</code>.</td> </tr> <tr> <td><code>psts</code></td> <td><code>(3, 48, n, m, 2)</code></td> <td>Known position of bounding box. First value is distance to left edge, second value distance to top edge.</td> </tr> <tr> <td><code>metadata</code></td> <td><code>(6, 2)</code></td> <td>Metadata on title, description, author, license, version and date.</td> </tr> <tr> <td><code>lbldata</code></td> <td><code>(n,)</code></td> <td>Descriptive names for each ground truth labels.</td> </tr> </tbody> </table> <p>For use in Python a dataset class is provided that implements the basic functionality for loading a certain split and scale selection, as illustrated in the code below. It ensures shuffling is done in a consistent manner such that ground truth scales and positions can be retrieved. Metadata and label descriptions can be retrieved via <code>metadata</code> and <code>labeldata</code>, respectively.</p> <pre><code class="language-python">from data.dataset import STIRDataset dataset = STIRDataset('data/emoji.npz') # Obtain images and labels for training images, labels = dataset.to_torch(split='train', scales=[32, 64], shuffle=True) # Obtain known scales and positions for above scales, positions = dataset.get_latents(split='train', scales=[32, 64], shuffle=True) # Get metadata and label descriptions metadata = dataset.metadata label_descriptions = dataset.labeldata</code></pre> <p><strong>License and Attribution</strong></p> <p>When using this dataset for your own research, please respect the individual licenses of the original data. These are distributed within&nbsp;the data files&#39; metadata. For attribution in papers, we recommend the following citations.</p> <ol> <li>D. Gandy, J. Otero, E. Emanuel, F. Botsford, J. Lundien, K. Jackson, M. Wilkerson, R. Madole, J. Raphael, T. Chase, G. Taglialatela, B. Talbot, and T. Chase. Font Awesome.&nbsp;https://fontawesome.com/v5/download, Nov. 2022.</li> <li>Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner. Gradient-based learning applied to document recognition. <em>Proc. IEEE</em>,&nbsp;86(11):2278&ndash;2324, Nov. 1998.</li> <li>&nbsp;C. Ertler, J. Mislej, T. Ollmann, L. Porzi, G. Neuhold, and Y. Kuang.&nbsp;The Mapillary Traffic Sign Dataset for Detection and Classification on a Global Scale. In <em>2020 16th Eur. Conf. Comput. Vision (ECCV)</em>, Glasgow, UK, Aug. 2020.</li> <li>G.-S.&nbsp;Xia, X.&nbsp;Bai, J.&nbsp;Ding, Z.&nbsp;Zhu, S.&nbsp;Belongie, J.&nbsp;Luo, M.&nbsp;Datcu, M.&nbsp;Pelillo, and L.&nbsp;Zhang. DOTA: A Large-Scale Dataset for Object Detection in Aerial Images. In <em>2018 IEEE/CVF Conf. Comput. Vision and Pattern Recognition (CVPR)</em>, pages 3974&ndash;3983, Salt Lake City, UT, USA, June 2018.</li> </ol>

opencc-by-4.0Nov 2022View details →
dryad40/100

Data from: The CellPhe toolkit for cell phenotyping using time-lapse imaging and pattern recognition

<p>With phenotypic heterogeneity in whole cell populations widely recognised, the demand for quantitative and temporal analysis approaches to characterise single cell morphology and dynamics has increased. We present CellPhe, a pattern recognition toolkit for the unbiased characterisation of cellular phenotypes within time-lapse videos. CellPhe imports tracking information from multiple segmentation and tracking algorithms to provide automated cell phenotyping from different imaging modalities, including fluorescence. To maximise data quality for downstream analysis, our toolkit includes automated recognition and removal of erroneous cell boundaries induced by inaccurate tracking and segmentation. We provide an extensive list of features extracted from individual cell time series, with custom feature selection to identify variables that provide the greatest discrimination for the analysis in question. Using ensemble classification for accurate prediction of cellular phenotype and clustering algorithms for the characterisation of heterogeneous subsets, we validate and prove adaptability using different cell types and experimental conditions.</p>

opencc-zeroFeb 2023View details →
zenodo40/100

An Occlusion and Pose Sensitive Image Dataset for Black Ear Recognition

<p><strong>RESEARCH APPROACH</strong></p> <p>The research approach adopted for the study consists of seven phases which includes as shown in Figure 1:</p> <ol> <li>Pre-acquisition</li> <li>data pre-processing</li> <li>Raw images collection</li> <li>Image pre-processing</li> <li>Naming of images</li> <li>Dataset&nbsp;Repository</li> <li>Performance Evaluation</li> </ol> <p>The different phases in the study are discussed in the sections below.</p> <p>&nbsp;</p> <p><strong>PRE-ACQUISITION</strong></p> <p>The volunteers are given brief orientation on how their data will be managed and used for research purposes only. After the volunteers agrees, a consent form is given to be read and signed. The sample of the consent form filled by the volunteers is shown in Figure 1.</p> <p>The capturing of images was started with the setup of the imaging device. The camera is set up on a tripod stand in stationary position at the height 90 from the floor and distance 20cm from the subject.</p> <p>&nbsp;</p> <p><strong>EAR </strong><strong>IMAGE ACQUISITION</strong></p> <p>Image acquisition is an action of retrieving image from an external source for further processing. The image acquisition is purely a hardware dependent process by capturing unprocessed images of the volunteers using a professional camera. This was acquired through a subject posing in front of the camera. It is also a process through which digital representation of a scene can be obtained. This representation is known as an image and its elements are called pixels (picture elements). The imaging sensor/camera used in this study is a Canon E0S 60D professional camera which is placed at a distance of 3 feet form the subject and 20m from the ground.&nbsp;</p> <p>This is the first step in this project to achieve the project&rsquo;s aim of developing an occlusion and pose sensitive image dataset for black ear recognition. (OPIB ear dataset). To achieve the objectives of this study, a set of black ear images were collected mostly from undergraduate students at a public University in Nigeria.</p> <p>&nbsp;</p> <p>The image dataset required is captured in two scenarios:</p> <p>1. uncontrolled environment with a surveillance camera</p> <ol> </ol> <p>The image dataset captured is purely black ear with partial occlusion in a constrained and unconstrained environment.</p> <p>&nbsp;</p> <p>2. controlled environment with professional cameras</p> <p>The ear images captured were from black subjects in controlled environment. To make the OPIB dataset pose invariant, the volunteers stand on a marked positions on the floor indicating the angles at which the imaging sensor was captured the volunteers&rsquo; ear. The capturing of the images in this category requires that the subject stand and rotates in the following angles 60<sup>o</sup>, 30<sup>o</sup> and 0<sup>o</sup> towards their right side to capture the left ear and then towards the left to capture the right ear (Fernando <em>et al.,</em> 2017) as shown in Figure 4. Six (6) images were captured per subject at angles 60<sup>o</sup>, 30<sup>o</sup> and 0<sup>o</sup> for the left and right ears of 152 volunteers making a total of 907 images <strong><em>(five volunteers had 5 images instead of 6, hence f</em></strong><strong><em>olders 34, 22, 51, 99 and&nbsp;102 contain 5 images).</em></strong></p> <p>To make the OPIB dataset occlusion and pose sensitive, partial occlusion of the subject&rsquo;s ears were simulated using rings, hearing aid, scarf, earphone/ear pods, etc. before the images are captured.</p> <p>&nbsp;</p> <table> <tbody> <tr> <td> <p><strong>CONSENT FORM</strong></p> <p>This form was designed to obtain participant&rsquo;s consent on the project titled: <strong>An Occlusion and Pose Sensitive Image Dataset for Black Ear Recognition</strong><strong> (OPIB)</strong>. The information is purely needed for academic research purposes and the ear images collected will curated anonymously and the identity of the volunteers will not be shared with anyone. The images will be uploaded on online repository to aid research in ear biometrics.</p> <p>The participation is voluntary, and the participant can withdraw from the project any time before the final dataset is curated and warehoused.</p> <p>Kindly sign the form to signify your consent.</p> <p><strong><em>I consent to my image being recorded in form of still images or video surveillance as part of the OPIB ear images project.</em></strong></p> <p><strong>Tick as appropriate:</strong></p> <p><strong>GENDER</strong> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Male &nbsp;&nbsp; Female</p> <p><strong>AGE</strong> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; (18-25)&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; (26-35)&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; (36-50)</p> <p>&nbsp;</p> <p>&hellip;&hellip;&hellip;&hellip;&hellip;&hellip;&hellip;&hellip;&hellip;&hellip;&hellip;&hellip;..</p> <p>SIGNED</p> </td> </tr> </tbody> </table> <p>&nbsp;</p> <p><strong>Figure 1</strong>: Sample of Subject&rsquo;s Consent Form for the OPIB ear dataset</p> <p>&nbsp;</p> <p><strong>RAW IMAGE COLLECTION</strong></p> <p>The ear images were captured using a digital camera which was set to JPEG because if the camera format is set to raw, no processing will be applied, hence the stored file will contain more tonal and colour data. However, if set to JPEG, the image data will be processed, compressed and stored in the appropriate folders.</p> <p>&nbsp;</p> <p><strong>IMAGE PRE-PROCESSING </strong></p> <p>The aim of pre-processing is to improve the quality of the images with regards to contrast, brightness and other metrics. It also includes operations such as: cropping, resizing, rescaling, etc. which are important aspect of image analysis aimed at dimensionality reduction. The images are downloaded on a laptop for processing using MATLAB.</p> <p>&nbsp;</p> <p><strong>Image Cropping</strong></p> <p>The first step in image pre-processing is image cropping. Some irrelevant parts of the image can be removed, and the image Region of Interest (ROI) is focused. This tool provides a user with the size information of the cropped image. MATLAB function for image cropping realizes this operation interactively by waiting for a user to specify the crop rectangle with the mouse and operate on the current axes. The output images of the cropping process are of the same class as the input image.</p> <p><strong>Naming of OPIB Ear Images</strong></p> <p>The OPIB ear images were labelled based on the naming convention formulated from this study as shown in Figure 5. The images are given unique names that specifies the subject, the side of the ear (left or right) and the angle of capture. The first and second letters (SU) in the image names is block letter simply representing subject for subject 1-to-n in the dataset, while the left and right ears is distinguished using L1, L2, L3 and R1, R2, R3 for angles 60<sup>0</sup>, 30<sup>0</sup> and 0<sup>0</sup><sub>, </sub>respectively as shown in Table 1.</p> <p>&nbsp;</p> <p><strong>Table 1: Naming Convention for OPIB ear images</strong></p> <table align="center"> <tbody> <tr> <td> <p>NAMING CONVENTION</p> </td> </tr> <tr> <td> <p>Label</p> <p>Degrees&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 60<sup>0</sup>&nbsp;&nbsp;&nbsp;&nbsp; 30<sup>0</sup>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 0<sup>0</sup></p> </td> </tr> <tr> <td> <p>No of the degree&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 3</p> </td> </tr> <tr> <td> <p>Subject 1&nbsp;&nbsp; indicates&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; (first image in dataset) SU<sub>1</sub></p> </td> </tr> <tr> <td> <p>Subject n&nbsp;&nbsp; indicates&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; (last image in dataset) SU<sub>n</sub></p> </td> </tr> <tr> <td> <p>Left Image 1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; L 1</p> <p>Left image n&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; L n</p> <p>Right Image 1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; R 1</p> <p>Right Image n&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; R n</p> </td> </tr> <tr> <td> <p>SU1L<sub>1</sub>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; SU1R<sub>I</sub></p> <p>SU1L<sub>2</sub>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; SU1R<sub>2</sub>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</p> <p>SU1L<sub>3</sub>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; SU1R<sub>3</sub></p> </td> </tr> </tbody> </table> <p>&nbsp;</p> <p><strong>OPIB EAR DATASET EVALUATION</strong></p> <p>The prominent challenges with the current evaluation practices in the field of ear biometrics are the use of different databases, different evaluation matrices, different classifiers that mask the feature extraction performance and the time spent developing framework (Abaza <em>et al.</em>, 2013; Emer&scaron;ič <em>et al.,</em> 2017).</p> <p>The toolbox provides environment in which the evaluation of methods for person recognition based on ear biometric data is simplified. It executes all the dataset reads and classification based on ear descriptors.</p> <p>&nbsp;</p> <p><strong>DESCRIPTION OF OPIB EAR DATASET</strong></p> <p>OPIB ear dataset was organised into a structure with each folder containing 6 images of the same person. The images were captured with both left and right ear at angle 0, 30 and 60 degrees. The images were occluded with earing, scarves and headphone etc. &nbsp;The collection of the dataset was done both indoor and outdoor.&nbsp; The dataset was gathered through the student at a public university in Nigeria. The percentage of female (40.35%) while Male (59.65%).&nbsp; The ear dataset was captured through a profession camera Nikon D 350. It was set-up with a camera stand where an individual captured in a process order. A total number of 907 images was gathered.</p> <p>The challenges encountered in term of gathering students for capturing, processing of the images and annotations. The volunteers were given a brief orientation on what their ear could be used for before, it was captured, for processing.&nbsp; It was a great task in arranging the ear (dataset) into folders and naming accordingly.</p> <p>&nbsp;</p> <p><strong>Table 2</strong>: Overview of the OPIB Ear Dataset</p> <table align="left"> <tbody> <tr> <td> <p>Location</p> </td> <td> <p>Both Indoor and outdoor environment</p> </td> </tr> <tr> <td> <p>Information about Volunteers</p> </td> <td> <p>Students</p> </td> </tr> <tr> <td> <p>Gender</p> </td> <td> <p>Female (40.35%) and male (59.65%)</p> </td> </tr> <tr> <td> <p>Head Side Left and Right</p> </td> <td> <p>Side Left and Right</p> </td> </tr> <tr> <td> <p>Total number of volunteers</p> </td> <td> <p>152</p> </td> </tr> <tr> <td> <p>Per Subject images</p> </td> <td> <p>3 images of left ear and 3 images of right ear</p> </td> </tr> <tr> <td> <p>Total Images</p> </td> <td> <p>907</p> </td> </tr> <tr> <td> <p>Age group</p> </td> <td> <p>18 to 35 years</p> </td> </tr> <tr> <td> <p>Colour Representation</p> </td> <td> <p>RGB</p> </td> </tr> <tr> <td> <p>Image Resolution</p> </td> <td> <p>224x224</p> </td> </tr> </tbody> </table> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p> <p>&nbsp;</p>

opencc-by-4.0Mar 2023View details →
zenodo40/100

Interpretation of EKG with Image Recognition and Convolutional Neural Networks

<p>Dataset used in the training and evaluation of &quot;Interpretation of EKG with Image Recognition and Convolutional Neural Networks&quot;</p> <p>Abstract:</p> <p>Electrocardiograms (EKG) form the backbone of all cardiovascular diagnosis, treatment and follow up. Given the pivotal role it plays in modern medicine, there have been multiple efforts to computerize the EKG interpretation with algorithms to improve efficiency and accuracy. Unfortunately, many of these algorithms are machine specific and run-on proprietary signals generated by that machine, hence not generalizable. We propose the development of an image recognition model which can be used to read standard EKG strips. A convolutional neural network (CNN) was trained to classify 12-lead EKGs between 7 clinically important diagnostic classes. An austere variation of the MobileNetV3 model was trained from the ground up on publicly available labeled training set. The precision per class varies from 52% to 91%. This is a novel approach to EKG interpretation as an image recognition problem.</p>

opencc-by-4.0May 2023View details →
dryad40/100

Data from: The CellPhe toolkit for cell phenotyping using time-lapse imaging and pattern recognition

Open the record for dataset details and reuse information.

publicFeb 2023View details →
zenodo36/100

Multi-sensor dataset for testing merge of Hyperspectral, HD and 3D cloud information for image recognition

<p>This data contain&nbsp;multisensor image dataset constructed for benchmarking purposes. It contains multiple images of the constructed scenes -- on which objects made of different materials are placed to test image recognition scenarios. The scene is recorded from various angles by imagining sensors, i.e. HSI camera, HD camera on mobile chassis and MS Kinect to provide complete information.<br> &nbsp;</p> <p><strong>Equipment</strong><br> The imaging was performed with use of three devices for three different approaches to data. Those three devices&#39; imaging characteristics are widely different when it comes to angle and resolution which required them to be separately positioned to acquire the matching images. Therefore while HD camera was being transferred on the moving platform (chassis), both Kinect and SOC710 were placed on a stationary position which was moved between the frames by hand.</p> <p><em>Hyperspectral data</em></p> <p><br> Hyperspectral data acquisition was performed with Surface Optics SOC710 camera. &nbsp;This camera records spectra at VNIR range $377-1046$ nm; the output image has dimensions $696 \times 520$ with 128 bands and $12$ bit dynamic range.</p> <p>The camera is equipped with sensor line translation unit and can be used from static stand as a conventional camera (i.e. it does not require mechanical translation of the observed sample or rotary stand, as in traditional &lsquo;push broom&rsquo; hyperspectral cameras). The lighting was provided with four ambient lamps and adjusted for each scenario separately, so that most of the dynamic range of the camera was used and image saturation is avoided. Captured hyperspectral images were subject to a standard calibration procedure, including: the removal of a dark frame, spectral and radiometric calibration as well as reflectance normalization using the calibration panel.&nbsp;</p> <p><em>3D point clouds</em></p> <p><br> The Kinect sensor incorporates several advanced sensing hardware. The depth sensor consists of the IR projector combined with the IR camera, which is a monochrome complementary metaloxide semiconductor (CMOS) sensor. The IR projector is an IR laser that passes through a diffraction grating and turns into a set of IR dots. The relative geometry between the IR projector and the IR camera as well as the projected IR dot pattern are known. If we can match a dot observed in an image with a dot in the projector pattern, we can reconstruct it in 3D using triangulation. Because the dot pattern is relatively random, the matching between the IR image and the projector pattern can be done in a straightforward way by comparing small neighborhoods using, for example, normalized cross correlation. The depth value is encoded with gray values; the darker a pixel, the closer the point is to the camera in space. The black pixels indicate that no depth values are available for those pixels. This might happen if the points are too far (and the depth values cannot be computed accurately), are too close (there is a blind region due to limited fields of view for the projector and the camera), are in the cast shadow of the projector (there are no IR dots), or reflect poor IR lights&nbsp;</p> <p><em>HD Images</em></p> <p><br> The HD images were acquired using 5 Megapixel HD camera mounted on a mobile chassis made by Dawn Robotics, that allowed the camera to be moved freely on the scene. Both camera and mobile chassis was controlled by a Raspberry PI unit which was also responsible to position the camera in accord to the data being collected by other sources.&nbsp;<br> &nbsp;</p> <p><strong>Data</strong></p> <p>The dataset consists of three scenes consisting of various objects -- minerals, fruit, wood plastic and metal -- placed on a stand. The objects, depending on the view are partially covered and seen from different perspective. Each scene is captured from 8 different angles.</p> <p>Scene 1 (denoted <em>SceneEagle</em>) uses mostly inorganic materials, such as wood, metal, plastic and glass all placed on the vertical stand.<br> Scene 2 (<em>SceneFruit</em>) uses fruits normal and artificial, that are similar on HD photography and 3D cloud of point, but differs in hyperspectral image.<br> Scene 3 (<em>SceneFruit2</em>) uses the fruits but also includes printed full colour images of same fruits that are 2-dimensional.</p> <p>&nbsp;</p> <p>The data are formatted as follows:<br> - The HIS images are available in both \text{*.hdr} and \text{*.cube} formats. The separate files with calibrating panel is provided for each frame.<br> -&nbsp; Kinect clouds are provided in \text{*.obj} format, typical for Kinect output files.<br> - Matched Hyperspectral clouds are also provided as \text{*.obj} files<br> - HD photo files are provided in \text{*.jpg} files.<br> &nbsp;</p> <p><br> <strong>Acknowledgements</strong></p> <p>This work has been supported by the National Science Centre, based on decision no. DEC2012/07/N/ST6/03656.</p> <p>&nbsp;</p>

opencc-by-4.0Aug 2020View details →

ScienceDex guides

Understand access before you commit

These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.

Compare curated datasets

Allen Brain Atlas

Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.

allen-brain-atlas
neuroscienceopenDocumentation, web resources, and API references are available online.
Last verified 2026-04-30Open record

Annotated Behaviour and Observability Dataset (ABODe)

ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.

abode-home-cage
behavioral-neuroscienceopenThe DataShare record exposes download links for annotations, documentation, license text, and the zipped per-snippet data directory.
Last verified 2026-04-30Open record

DANDI Archive for NWB datasets

DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.

dandi-nwb
electrophysiologyopenPublished Dandiset metadata and archive endpoints are available through the production DANDI API.
Last verified 2026-04-30Open record

International Brain Laboratory public data

The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.

ibl
behavioral-neuroscienceopenPublic sessions can be searched and loaded from the IBL public data server through ONE.
Last verified 2026-04-29Open record

OpenNeuro

OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.

openneuro
neuroscienceopenPublished datasets are available on demand over the internet.
Last verified 2026-04-29Open record