Dataset Restricted Access
2014 ImageCLEF WEBUPV Collection
Villegas, Mauricio; Paredes, Roberto
JSON Export
{
"owners": [
19451
],
"doi": "10.5281/zenodo.259758",
"stats": {
"version_unique_downloads": 13.0,
"unique_views": 263.0,
"views": 288.0,
"version_views": 288.0,
"unique_downloads": 13.0,
"version_unique_views": 263.0,
"volume": 110167422414.0,
"version_downloads": 62.0,
"downloads": 62.0,
"version_volume": 110167422414.0
},
"links": {
"latest_html": "https://zenodo.org/record/259758",
"doi": "https://doi.org/10.5281/zenodo.259758",
"badge": "https://zenodo.org/badge/doi/10.5281/zenodo.259758.svg",
"html": "https://zenodo.org/record/259758",
"latest": "https://zenodo.org/api/records/259758"
},
"created": "2017-01-25T17:52:13.421037+00:00",
"updated": "2020-01-24T19:21:42.466403+00:00",
"conceptrecid": "748690",
"revision": 7,
"id": 259758,
"metadata": {
"access_right_category": "danger",
"doi": "10.5281/zenodo.259758",
"description": "<p>This document describes the WEBUPV dataset compiled for the ImageCLEF 2014<br>\nScalable Concept Image Annotation challenge. The data mentioned here indicates what<br>\nis ready for download. However, upon request or depending on feedback from the<br>\nparticipants, additional data may be released.</p>\n\n<p>The following is the directory structure of the collection, and bellow there<br>\nis a brief description of what each compressed file contains. The<br>\ncorresponding MD5 checksums of the files shown (for verifying a correct<br>\ndownload) can be found in md5sums.txt.</p>\n\n<p>Directory structure<br>\n-------------------</p>\n\n<p>.<br>\n|<br>\n|--- README.txt<br>\n|--- md5sums.txt<br>\n|--- webupv14_train_lists.zip<br>\n|--- webupv14_train2_lists.zip<br>\n|--- webupv14_devel_lists.zip<br>\n|--- webupv14_test_lists.zip<br>\n|--- webupv14_baseline.zip<br>\n|<br>\n|--- feats_textual/<br>\n| |<br>\n| |--- webupv14_{train|train2}_textual_pages.zip<br>\n| |--- webupv14_{train|train2}_textual.scofeat.gz<br>\n| |--- webupv14_{train|train2}_textual.keywords.gz<br>\n|<br>\n|--- feats_visual/<br>\n |<br>\n |--- webupv14_{train|train2|devel|test}_visual_images.zip<br>\n |--- webupv14_{train|train2|devel|test}_visual_gist2.feat.gz<br>\n |--- webupv14_{train|train2|devel|test}_visual_sift_1000.feat.gz<br>\n |--- webupv14_{train|train2|devel|test}_visual_csift_1000.feat.gz<br>\n |--- webupv14_{train|train2|devel|test}_visual_rgbsift_1000.feat.gz<br>\n |--- webupv14_{train|train2|devel|test}_visual_opponentsift_1000.feat.gz<br>\n |--- webupv14_{train|train2|devel|test}_visual_colorhist.feat.gz<br>\n |--- webupv14_{train|train2|devel|test}_visual_getlf.feat.gz</p>\n\n<p><br>\nContents of files<br>\n-----------------</p>\n\n<p>* webupv14_train{|2}_lists.zip</p>\n\n<p> The first training set ("train_*") includes images for the concepts of the<br>\n development set, whereas the second training set ("train2_*") includes<br>\n images for the concepts in the test set that are not in the development set.</p>\n\n<p> -> train{|2}_iids.txt : IDs of the images (IIDs) in the training set.</p>\n\n<p> -> train{|2}_rids.txt : IDs of the webpages (RIDs) in the training set.</p>\n\n<p> -> train{|2}_*urls.txt : The original URLs from where the images (iurls)<br>\n and the webpages (rurls) were downloaded. Each line in the file<br>\n corresponds to an image, starting with the IID and is followed<br>\n by one or more URLs.</p>\n\n<p> -> train{|2}_rimgsrc.txt : The URLs of the images as referenced in each<br>\n of the webpages. Each line of the file is of the form: IID RID<br>\n URL1 [URL2 ...]. This information is necessary to locate the<br>\n images within the webpages and it can also be useful as a<br>\n textual feature.</p>\n\n<p>* webupv14_{devel|test}_lists.zip</p>\n\n<p> -> {devel|test}_conceptlists.txt : Lists per image of concepts for<br>\n annotation. Each line starts with an image ID and is followed by the<br>\n list of concepts in alphabetical order. Each ID may appear more than<br>\n once. In total there are 1940 image annotation lists for the<br>\n development set and 7291 image annotation lists for the test set. These<br>\n correspond to 1000 and 4122 unique images (IDs) for the development and<br>\n test sets, respectively.</p>\n\n<p> -> {devel|test}_allconcepts.txt : Complete list of concepts for the<br>\n development/test set.</p>\n\n<p> The concepts are defined by one or more WordNet synsets, which is<br>\n intended to make it possible to easily obtain more information about<br>\n the concepts, e.g. synonyms. In the concept list, the first column<br>\n (which is the name of the concept) indicates the word to search in<br>\n WordNet, the second column the synset type (either noun or adjective),<br>\n the third column is the sense number and the fourth column is the<br>\n WordNet offset (although this cannot be trusted since it changes<br>\n between WordNet versions). For most of the concepts there is a fifth<br>\n column which is a Wikipedia article related to the concept.</p>\n\n<p> -> {devel|test}_groundtruth.txt : Ground truth concepts for the development<br>\n and test sets.</p>\n\n<p> -> {devel|test}_*urls.txt : The original URLs from where the images (iurls)<br>\n and the webpages (rurls) were downloaded. Each line in the file<br>\n corresponds to an image, starting with the IID and is followed by one<br>\n or more URLs.</p>\n\n<p> Note: These are included only to acknowledge the source of the<br>\n data, not be used as input to the annotation systems.</p>\n\n<p><br>\n* webupv14_baseline.zip</p>\n\n<p> An archive that includes code for computing the evaluation measures<br>\n for two baseline techniques. See the included README.txt for<br>\n details.</p>\n\n<p><br>\n* feats_textual/webupv14_train{|2}_textual_pages.zip</p>\n\n<p> Contains all of the webpages which referenced the images in the<br>\n training set after being converted to valid xml. In total there are<br>\n 262588 files, since each image can appear in more than one page, and<br>\n there can be several versions of same page which differ by the<br>\n method of conversion to xml. To avoid having too many files in a<br>\n single directory (which is an issue for some types of partitions),<br>\n the files are found in subdirectories named using the first two<br>\n characters of the RID, thus the paths of the files after extraction<br>\n are of the form:</p>\n\n<p> ./WEBUPV/pages/{RID:0:2}/{RID}.{CONVM}.xml.gz</p>\n\n<p> To be able to locate the training images withing the webpages, the<br>\n URLs of the images as referenced are provided in the file<br>\n train_rimgsrc.txt.</p>\n\n<p>* feats_textual/webupv14_train{|2}_textual.scofeat.gz</p>\n\n<p> The processed text extracted from the webpages near where the images<br>\n appeared. Each line corresponds to one image, having the same order<br>\n as the train_iids.txt list. The lines start with the image ID,<br>\n followed by the number of extracted unique words and the<br>\n corresponding word-score pairs. The scores were derived taking into<br>\n account 1) the term frequency (TF), 2) the document object model<br>\n (DOM) attributes, and 3) the word distance to the image. The scores<br>\n are all integers and for each image the sum of scores is always<br>\n <=100000 (i.e. it is normalized).</p>\n\n<p><br>\n* feats_textual/webupv14_train{|2}_textual.keywords.gz</p>\n\n<p> The words used to find the images when querying image search<br>\n engines. Each line corresponds to an image (in the same order as in<br>\n train_iids.txt). The lines are composed of triplets:</p>\n\n<p> [keyword] [rank] [search_engine]</p>\n\n<p> where [keyword] is the word used to find the image, [rank] is the<br>\n position given to the image in the query, and [search_engine] is a<br>\n single character indicating in which search engine it was found<br>\n ('g':google, 'b':bing, 'y':yahoo).</p>\n\n<p><br>\n* feats_visual/webupv14_*_images.zip</p>\n\n<p> Contains thumbnails (maximum 640 pixels of either width or height)<br>\n of the images in jpeg format. To avoid having too many files in a<br>\n single directory (which is an issue for some types of partitions),<br>\n the files are found in subdirectories named using the first two<br>\n characters of the IID, thus the paths of the files after extraction<br>\n are of the form:</p>\n\n<p> ./WEBUPV/images/{IID:0:2}/{IID}.jpg</p>\n\n<p>* feats_visual/webupv14_*.feat.gz</p>\n\n<p> The visual features in a simple ASCII text sparse format. The first<br>\n line of the file indicates the number of vectors (N) and the<br>\n dimensionality (DIMS). Then each line corresponds to one vector,<br>\n starting with the number of non-zero elements and followed by pairs<br>\n of dimension-value, being the first dimension 0. In summary the file<br>\n format is:</p>\n\n<p> N DIMS<br>\n nz1 Dim(1,1) Val(1,1) ... Dim(1,nz1) Val(1,nz1)<br>\n nz2 Dim(2,1) Val(2,1) ... Dim(2,nz2) Val(2,nz2)<br>\n ...<br>\n nzN Dim(N,1) Val(N,1) ... Dim(N,nzN) Val(N,nzN)</p>\n\n<p> The order of the features is the same as in the lists<br>\n devel_conceptlists.txt, test_conceptlists.txt and train_iids.txt.</p>\n\n<p> The procedure to extract the SIFT based features in this<br>\n subdirectory was conducted as follows. Using the ImageMagick<br>\n software, the images were first rescaled to having a maximum of 240<br>\n pixels, of both width and height, while preserving the original<br>\n aspect ratio, employing the command:</p>\n\n<p> convert {IMGIN}.jpg -resize '240>x240>' {IMGOUT}.jpg</p>\n\n<p> Then the SIFT features where extracted using the ColorDescriptor<br>\n software from Koen van de Sande<br>\n (http://koen.me/research/colordescriptors). As configuration we<br>\n used, 'densesampling' detector with default parameters, and a hard<br>\n assignment codebook using a spatial pyramid as<br>\n 'pyramid-1x1-2x2'. The number in the file name indicates the size of<br>\n the codebook. All of the vectors of the spatial pyramid are given in<br>\n the same line, thus keeping only the first 1/5th of the dimensions<br>\n would be like not using the spatial pyramid. The codebook was<br>\n generated using 1.25 million randomly selected features and the<br>\n k-means algorithm. The GIST features were extracted using the<br>\n LabelMe Toolbox. The images where first resized to 256x256 ignoring<br>\n original aspect ratio, using 5 scales, 6 orientations and 4<br>\n blocks. The other features colorhist and getlf, are both color<br>\n histogram based extracted using our own implementation.</p>\n\n<p> </p>",
"title": "2014 ImageCLEF WEBUPV Collection",
"relations": {
"version": [
{
"count": 1,
"index": 0,
"parent": {
"pid_type": "recid",
"pid_value": "748690"
},
"is_last": true,
"last_child": {
"pid_type": "recid",
"pid_value": "259758"
}
}
]
},
"access_conditions": "<p>This dataset is available under a Creative Commons Attribution-<br>\nNonCommercial-ShareAlike 3.0 Unported License. Before downloading<br>\nthe data, please read and accept the Creative Commons License and<br>\nthe following usage agreement:</p>\n\n<p>Data Usage Agreement ImageCLEF 2012/2013/2014/2015/2016 WEBUPV Image<br>\nAnnotation Datasets</p>\n\n<p>By downloading the "Dataset", you (the "Researcher") agrees to the<br>\nfollowing terms.</p>\n\n<p>* The Researcher will only use the Dataset for non-commercial<br>\nresearch and/or educational purposes.</p>\n\n<p>* The Researcher will cite one of the following papers in any<br>\npublication that makes use of the Dataset.</p>\n\n<p> Gilbert, A., Piras, L., Wang, J., Yan, F., Ramisa, A.,<br>\n Dellandrea, E., Gaizauskas, R., Villegas, M., Mikolajczyk, K.:<br>\n Overview of the ImageCLEF 2016 scalable concept image<br>\n annotation task. In: CLEF2016 Working Notes, CEUR Workshop<br>\n Proceedings, CEUR-WS.org, Évora, Portugal, 5–8 September 2016</p>\n\n<p> Gilbert, A., Piras, L., Wang, J., Yan, F., Dellandrea, E.,<br>\n Gaizauskas, R., Villegas, M., Mikolajczyk, K.: Overview of the<br>\n ImageCLEF 2015 Scalable Image Annotation, Localization and<br>\n Sentence Generation task. In: CLEF2015 Working Notes. CEUR<br>\n Workshop Proceedings, CEUR-WS.org, Toulouse, France (September<br>\n 8-11 2015)</p>\n\n<p> Villegas, M., Paredes, R.: Overview of the ImageCLEF 2014<br>\n Scalable Concept Image Annotation Task. In: CLEF2014 Working<br>\n Notes. CEUR Workshop Proceedings, vol. 1180, pp. 308–328.<br>\n CEUR-WS.org, Sheffield, UK (September 15-18 2014)</p>\n\n<p> Villegas, M., Paredes, R., Thomee, B.: Overview of the ImageCLEF<br>\n 2013 Scalable Concept Image Annotation Subtask. In: CLEF 2013<br>\n Evaluation Labs and Workshop, Online Working Notes. Valencia,<br>\n Spain (September 23-26 2013)</p>\n\n<p>* The Researcher may provide research associates and colleagues a<br>\ncopy of the Dataset provided that they also agree to this Data<br>\nUsage Agreement.</p>\n\n<p>* The Researcher will assume all responsibility against any claims<br>\narising from Researcher's use of the Dataset.</p>",
"grants": [
{
"code": "600707",
"links": {
"self": "https://zenodo.org/api/grants/10.13039/501100000780::600707"
},
"title": "tranScriptorium",
"acronym": "TRANSCRIPTORIUM",
"program": "FP7",
"funder": {
"doi": "10.13039/501100000780",
"acronyms": [],
"name": "European Commission",
"links": {
"self": "https://zenodo.org/api/funders/10.13039/501100000780"
}
}
}
],
"communities": [
{
"id": "ecfunded"
},
{
"id": "imageclef"
}
],
"publication_date": "2014-04-01",
"creators": [
{
"affiliation": "Universitat Politecnica de Valencia",
"name": "Villegas, Mauricio"
},
{
"affiliation": "Universitat Politecnica de Valencia",
"name": "Paredes, Roberto"
}
],
"meeting": {
"acronym": "CLEF",
"url": "http://clef2014.clef-initiative.eu/",
"dates": "15-18 September 2014",
"place": "Sheffield, UK",
"title": "Conference and Labs of the Evaluation Forum"
},
"access_right": "restricted",
"resource_type": {
"type": "dataset",
"title": "Dataset"
},
"related_identifiers": [
{
"scheme": "url",
"identifier": "http://ceur-ws.org/Vol-1180/CLEF2014wn-Image-VillegasEt2014.pdf",
"relation": "isSupplementTo"
},
{
"scheme": "url",
"identifier": "http://imageclef.org/2014/annotation",
"relation": "isSupplementTo"
}
]
}
}
288
62
views
downloads
| All versions | This version | |
|---|---|---|
| Views | 288 | 288 |
| Downloads | 62 | 62 |
| Data volume | 110.2 GB | 110.2 GB |
| Unique views | 263 | 263 |
| Unique downloads | 13 | 13 |
Indexed in
- Publication date:
- April 1, 2014
- DOI:
-
- Grants:
-
European Commission:
- TRANSCRIPTORIUM - tranScriptorium (600707)
- Meeting:
- Conference and Labs of the Evaluation Forum (CLEF), Sheffield, UK, 15-18 September 2014
- Related identifiers:
- Supplement to
- Communities: