{
  "fetched_at": "2026-10-19T18:55:29Z",
  "source_url": "https://www.kaggle.com/competitions/awic2012/data",
  "pages": [
    {
      "name": "Description",
      "content": "<p>Writer identification is a very active research field. It is of a primordial importance in forensic document examination when it helps experts in delibirating on the authenticity of a certain document.</p>\r\n<p>This is a follow-up contest of the last year' <a href=\"http://www.kaggle.com/c/wic2011\" target=\"_blank\">\r\nArabic Writer Identification Contest</a>.</p>\r\n<p>As we mentioned in the last edition,&nbsp;<span>writer identification generally requires two steps. The first one is an image-processing step, where features are extracted from the images. The second step is a classification step, where the document is assigned\r\n to the “closest” document in the dataset according to the “difference” between their features.</span></p>\r\n<p><span>With regards to the last edition, this contest has several improvments:</span></p>\r\n<ul>\r\n<li>we have significantly augmented the number of writers (we have more than 200 writers in this new database).\r\n</li><li>we will not be providing any side-information (eg. number of documents per writer), as this is not necessarly known in a real forensic casework.\r\n</li><li>we will only provide binary images, as color and gray-level images might transform this into a pen identification task.\r\n</li></ul>\r\n<p>This competition is organized in conjunction with the International Conference of Frontiers in Handwriting Recognition\r\n<a href=\"http://www.icfhr2012.uniba.it/\" target=\"_blank\">ICFHR2012</a> which will be held in Bari, Italy in September&nbsp;18-20.</p>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Rules",
      "content": "\r\n\t<!-- Begin Base Rules -->\r\n    <h3>One account per participant</h3>\r\n    <p>You cannot sign up to Kaggle from multiple accounts and therefore you cannot submit from multiple accounts.</p>\r\n    <h3>No private sharing outside teams</h3>\r\n    <p>\r\n        Privately sharing code or data outside of teams is not permitted.\r\n        It's okay to share code if made available to all participants on the forums.\r\n    </p>\r\n    <h3>Public dissemination of entries</h3>\r\n    <p>Kaggle and the competition host have the right to publicly disseminate any entries or models.</p>\r\n    <h3>Open licensing of winners</h3>\r\n    <p>\r\n        Winning solutions need to be made available under\r\n        <a href=\"http://opensource.org/licenses\">a popular OSI-approved license</a>\r\n        in order to be eligible for recognition and prize money.\r\n    </p>\r\n    <h3>Winning solutions must be posted or linked to in the forums.</h3>\r\n    <p>\r\n        Prizes will be awarded after the winners have posted their solutions to the competition forum.\r\n        Winners must post or link to their solutions within seven (7) days of being notified of their winning status.\r\n    </p>\r\n    <h3>Team Mergers</h3>\r\n    <p>\r\n        Team mergers are allowed and can be performed by the team leader. In order to merge, the combined team must have a total submission count less than or equal to the maximum allowed as of the merge date. The maximum allowed is the number of submissions per day multiplied by the number of days the competition has been running.<br/>\r\n    </p>\r\n\r\n    <h3>Team Limits</h3>\r\n    <p>There is no maximum team size.</p>\r\n    <h3>Submission Limits</h3>\r\n    <p>You may submit a maximum of 2 entries per day.</p>\r\n    <p>You may select up to 3 final submissions for judging.</p>\r\n    <h3>Competition Timeline</h3>\r\n    <p>Start Date: <strong>2/21/2012 11:07 PM UTC</strong></p>\r\n    <p>Merger Deadline: <strong>None</strong></p>\r\n    <p>Entry Deadline: <strong>None</strong></p>\r\n    <p>End Date: <strong>4/15/2012 11:59 PM UTC</strong></p>\r\n\t<!-- End Base Rules -->\r\n<ol>\r\n<li>Winning participants are required to provide code as well as a description of their method.\r\n</li><li>An article about this contest will be published in the <a href=\"http://www.icfhr2012.uniba.it\" target=\"_blank\">\r\nICFHR2012</a> proceedings, by making a submission, you will be asked to provide a short description of your method to be included in this article.\r\n</li><li>Participants are free to use the provided features or to implement their own features. The use of manually extracted features or any other external data is not permitted.\r\n</li><li>The use of the competition data&nbsp;is not permitted for commercial purposes. </li><li>The use of&nbsp;the competition data&nbsp;is permitted for research purposes with the condition of properly citing the article mentioned above in any related publication.\r\n</li></ol>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Prizes",
      "content": "<p>The winner of this competition will be awarded $1000 and will be <span>acknowledged&nbsp;</span>in a special session at the&nbsp;<a href=\"http://www.icfhr2012.uniba.it\" target=\"_blank\">ICFHR2012</a>&nbsp;conference.</p>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "data-description",
      "content": "<p><a href=\"https://github.com/benhamner/awic2012\">Additional 1-Nearest-Neighbor benchmark in Python</a></p>\r\n<p><span>In this contest, more than 200 writers were asked to write three different paragraphs in Arabic language.&nbsp;The first two paragraphs are used for training and the third one for testing. For some writers, the first two paragraphs have been removed from\r\n the training set to test the ability of systems to detect unknown writers. Also, some writers have written the third paragraph more than once and some other participants did not writer the third paragraph at all.</span></p>\r\n<div>\r\n<div>Images are provided in PNG binary format. The binarization has been performed using Otsu's method. The images are provided in two subfolders “train” and “test”.&nbsp;The folder “train” contains images having the following format XXX_Y.png where XXX represents\r\n the ID of the writer and Y represents the number of the paragraph.</div>\r\n<div>The folder “test” contains images of the third paragraph, they all have the following format ZZ.png, where ZZ is the ID of the image in the test set.</div>\r\n<div><br>\r\nParticipants are asked to provide for each ZZ image, the ID of the most probable writer (among those which are in the training set). Like mentioned before, some ZZ images do not actually correspond to any writer in the training set. Participants are supposed\r\n to produce an ID eqal to 0 for those images.</div>\r\n<div>The competition is judged using the caterogization accuracy, which corresponds to the percentage of correctly identified writers.</div>\r\n<div>In the event there is a tie on the private leaderboard, the tie will be broken by the submission time (the prize money will go to the first submission).</div>\r\n<div>For participants who are not familiar with image processing, some geometric features are provided. These features are given in the form of histograms. The list below shows the name and the number of values in each of these features:</div>\r\n<div>\r\n<ul>\r\n<li>LengthsOfBranchesHist_10 (10 values) </li><li>ThicknessLengthsCircleHist30&nbsp;(30 values) </li><li>tortuosityHist10&nbsp;(10 values) </li><li>tortuosityDirectionHist10&nbsp;(10 values) </li><li>tortuosityDerivateHist10&nbsp;(10 values) </li><li>tortuosityDerivateDirectionHist10&nbsp;(10 values) </li><li>DirectionPerpendicular5Hist10&nbsp;(10 values) </li><li>CurvaturePerpendicular5Hist100&nbsp;(100 values) </li><li>CurvatureAli5Hist100&nbsp;(100 values) </li><li>CurvaturesDerivateAli5Hist100&nbsp;(100 values) </li><li>CurvatureAli10Hist100&nbsp;(100 values) </li><li>CurvaturesDerivateAli10Hist100&nbsp;(100 values) </li><li>CurvatureAli15Hist100&nbsp;(100 values) </li><li>CurvaturesDerivateAli15Hist100&nbsp;(100 values) </li><li>CurvatureAli20Hist100&nbsp;(100 values) </li><li>CurvaturesDerivateAli20Hist100&nbsp;(100 values) </li><li>chaincodeHist_4&nbsp;(4 values) </li><li>chaincodeHist_8&nbsp;(8 values) </li><li>chaincode8order2_64&nbsp;(64 values) </li><li>chaincode4order2_16&nbsp;(16 values) </li><li>chaincode4order3_64&nbsp;(64 values) </li><li>chaincode8order3_512&nbsp;(512 values) </li><li>chaincode4order4_256&nbsp;(256 values) </li><li>chaincode8order4_4096&nbsp;(4096 values) </li><li>distribution_types_73&nbsp;(73 values) </li><li>distribution_types_differences_73&nbsp;(73 values) </li><li>directions_hist1_4&nbsp;(4 values) </li><li>directions_hist2_8&nbsp;(8 values) </li><li>directions_hist3_12&nbsp;(12 values) </li><li>directions_hist4_16&nbsp;(16 values) </li><li>directions_hist1a2_12&nbsp;(12 values) </li><li>directions_hist1a2a3_24&nbsp;(24 values) </li><li>directions_hist1a2a3a4_40&nbsp;(40 values) </li></ul>\r\n</div>\r\n<div>Participants are free to use these features or their own features or even a combination of both.<br>\r\nThe submissions must be in a 2 columns format: The first one contains in ID of the questioned document and the second one, the ID of the most probable writer (or 0 in case of unknown writer). Please see&nbsp;<a href=\"http://www.kaggle.com/c/awic2012/download/sample_entry.csv\">sample_entry.csv</a>&nbsp;for\r\n an example.</div>\r\n<div>Finally, a sample code implementing the Edge-Based Directional Features in matlab is provided along with the corresponding benchmarks (cf.\r\n<a href=\"http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=1227797\">Bulacu et al. 2003' article</a><span class=\"x_x_x_x_f\"> for further details about the method</span>).</div>\r\n</div>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Custom",
      "content": "<p><span>The dataset used for the competition is a sub-set of a large dataset that has been acquired in Qatar University and will soon be made available. The data collection has been done by&nbsp;<span>the research assistant Wael Al-Ayoubi within a research project\r\n under the supervision of&nbsp;<a class=\"x_x_x_x_x_x_x_x_x_x_x_x_x_x_x_x_x_external\" href=\"http://faculty.qu.edu.qa/salali/\">Dr. Somaya Al-Maadeed</a><span>.</span></span></span></p>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}