{
  "fetched_at": "2026-10-24T04:02:10Z",
  "source_url": "https://www.kaggle.com/competitions/danburyaijuneworkshop/data",
  "pages": [
    {
      "name": "rules",
      "content": "Don't cheat. ",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "data-description",
      "content": "<h2>The Street View House Numbers (SVHN) Dataset</h2>\nSVHN is a real-world image dataset for developing machine learning and object recognition algorithms with minimal requirement on data preprocessing and formatting. It can be seen as similar in flavor to MNIST (e.g., the images are of small cropped digits), but comes from a significantly harder, unsolved, real world problem (recognizing digits and numbers in natural scene images). SVHN is obtained from house numbers in Google Street View images.\n\n* 10 classes, 1 for each digit. Digit '1' has label 1, '9' has label 9 and '0' has label 10.\n* MNIST-like 32-by-32 images centered around a single character (many of the images do contain some distractors at the sides).\n\n<h2>File descriptions</h2>\n<ul>\n<li><strong>train_x.npy</strong> - the training set features as a numpy matrix. Dimension: (43954, 32, 32, 3).</li>\n<li><strong>train_y.npy</strong> - the training set labels as a numpy matrix. Dimension: (43954,).</li>\n<li><strong>test.npy</strong> - the test set as a numpy matrix. Dimension: (29303, 32, 32, 3)</li>\n<li><strong>example_submission.csv</strong> - a sample submission file in the correct format</li>\n</ul>\nWe are dealing with RGB images in this competition. The dimensions of the training set features matrix should be read as: \n43954 rgb images, each of which are 32 by 32 and have 3 channels corresponding to r,g,b. ",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "# DanburyAI: June 2018 Workshop Competition\n## The Street View House Numbers (SVHN) Dataset\nSVHN is a real-world image dataset for developing machine learning and object recognition algorithms with minimal requirement on data preprocessing and formatting. It can be seen as similar in flavor to MNIST (e.g., the images are of small cropped digits), but comes from a significantly harder, unsolved, real world problem (recognizing digits and numbers in natural scene images). SVHN is obtained from house numbers in Google Street View images. \n\n - 10 classes, 1 for each digit. Digit '1' has label 1, '9' has label 9 and '0' has label 10.\n - MNIST-like 32-by-32 images centered around a single character (many of the images do contain some distractors at the sides).\n\nHere is a selection of what the individual training images look like: \n![enter image description here][1]\n\n## Source\nYuval Netzer, Tao Wang, Adam Coates, Alessandro Bissacco, Bo Wu, Andrew Y. Ng Reading Digits in Natural Images with Unsupervised Feature Learning NIPS Workshop on Deep Learning and Unsupervised Feature Learning 2011. ([PDF][2])\n\nWeb source: [http://ufldl.stanford.edu/housenumbers/][3]\n\n\n  [1]: http://ufldl.stanford.edu/housenumbers/32x32eg.png\n  [2]: http://ufldl.stanford.edu/housenumbers/nips2011_housenumbers.pdf\n  [3]: http://ufldl.stanford.edu/housenumbers/",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "For this competition, we will use multi-class classification accuracy:\n<p>\\[\\large\\frac{1}{n}\\sum_i^n{\\hat{y_i} == y_i}\\] </p>\nWhere:\n\n - y hat is the ground truth label and y is the predicted label.\n - n is the number of images in the test set. \n - The double equals sign represents a boolean function which returns 1 if the predicted label is equal to the actual label and 0 otherwise. ",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}