{
  "total": 100,
  "topics": [
    {
      "id": 602572,
      "title": "system & probe used (reproducibility)",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 316542,
      "title": "What about RLE value for \"No Mask\"?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 300972,
      "title": "Generate a DOI for uploaded datasets.",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 239762,
      "title": "Dataset information",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 116481,
      "title": "~0.69 easy-customisable U-net, Keras",
      "comment_count": 9,
      "views": 0,
      "votes": 8
    },
    {
      "id": 175258,
      "title": "Invalid file path",
      "comment_count": 2,
      "views": 0,
      "votes": 0
    },
    {
      "id": 112463,
      "title": "Is the challenge still work?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 41114,
      "title": "Cannot download the data",
      "comment_count": 10,
      "views": 0,
      "votes": 1
    },
    {
      "id": 21211,
      "title": "Permission to use data ",
      "comment_count": 10,
      "views": 1265,
      "votes": 7
    },
    {
      "id": 22982,
      "title": "Leaderboard Results Final!",
      "comment_count": 3,
      "views": 392,
      "votes": 4
    },
    {
      "id": 22951,
      "title": "Would like to see how others approached this problem. Doesn't have to be top LB.",
      "comment_count": 40,
      "views": 4165,
      "votes": 4
    },
    {
      "id": 22546,
      "title": "~0.62 Torch(net) Tutorial + Code",
      "comment_count": 36,
      "views": 6487,
      "votes": 21
    },
    {
      "id": 21358,
      "title": "~0.57 Deep Learning Keras Tutorial",
      "comment_count": 88,
      "views": 25058,
      "votes": 82
    },
    {
      "id": 69009,
      "title": "Viability of Application",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 61733,
      "title": "Data Usage",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 55362,
      "title": "Use of data",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 46922,
      "title": "How can evaluate the results?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 43987,
      "title": "Will the late submission result appear on the leader board?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 40499,
      "title": "Need a little help in Final Year Project",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 33675,
      "title": "what does it mean the rows in train_masks.csv.zip file",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 29872,
      "title": "Using the data",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 28881,
      "title": "Regarding usage of dataset- Ultrasound Nerve Segmentation",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 28062,
      "title": "Using the Data for a Science Fair Project",
      "comment_count": 0,
      "views": 35,
      "votes": 1
    },
    {
      "id": 22649,
      "title": "Manually editing trainset",
      "comment_count": 3,
      "views": 515,
      "votes": 1
    },
    {
      "id": 22905,
      "title": "When the competition is done,  I would appreciate for people willing to share their experience here!",
      "comment_count": 24,
      "views": 5832,
      "votes": 8
    },
    {
      "id": 22958,
      "title": "~0.7 Keras code",
      "comment_count": 17,
      "views": 2768,
      "votes": 24
    },
    {
      "id": 23136,
      "title": "LSUV initialization script for Keras ",
      "comment_count": 0,
      "views": 476,
      "votes": 4
    },
    {
      "id": 22985,
      "title": "Here are my thoughts (25th private - 0.70679)",
      "comment_count": 9,
      "views": 1590,
      "votes": 24
    },
    {
      "id": 23094,
      "title": "Code + ideas for 57th solution (0.691)",
      "comment_count": 0,
      "views": 653,
      "votes": 11
    },
    {
      "id": 23006,
      "title": "ImageNet segmentation 2016",
      "comment_count": 6,
      "views": 977,
      "votes": 2
    },
    {
      "id": 23009,
      "title": "Interested in seeing LB scores with % of images with mask predicted",
      "comment_count": 5,
      "views": 643,
      "votes": 2
    },
    {
      "id": 22953,
      "title": "If we all had a time machine!",
      "comment_count": 3,
      "views": 645,
      "votes": 2
    },
    {
      "id": 23004,
      "title": "Models for the existence of the mask?",
      "comment_count": 2,
      "views": 367,
      "votes": null
    },
    {
      "id": 22967,
      "title": "hardware",
      "comment_count": 5,
      "views": 470,
      "votes": null
    },
    {
      "id": 22955,
      "title": "your advice is appreciated",
      "comment_count": 1,
      "views": 291,
      "votes": 4
    },
    {
      "id": 22944,
      "title": "Since we are all waiting ...",
      "comment_count": 6,
      "views": 587,
      "votes": null
    },
    {
      "id": 22979,
      "title": "How to do it with R? ",
      "comment_count": 0,
      "views": 235,
      "votes": 0
    },
    {
      "id": 22950,
      "title": "What now?",
      "comment_count": 0,
      "views": 230,
      "votes": 0
    },
    {
      "id": 22887,
      "title": "Leaderboard Shakeup? What are the chances? ",
      "comment_count": 9,
      "views": 798,
      "votes": null
    },
    {
      "id": 22876,
      "title": "Looking for partners",
      "comment_count": 3,
      "views": 429,
      "votes": null
    },
    {
      "id": 22897,
      "title": "I'm sticking with my \"empty submission benchmark\"",
      "comment_count": 4,
      "views": 657,
      "votes": -12
    },
    {
      "id": 22875,
      "title": "CNN prediction doesn't change",
      "comment_count": 0,
      "views": 209,
      "votes": 0
    },
    {
      "id": 22747,
      "title": "Single model performance?",
      "comment_count": 29,
      "views": 3505,
      "votes": 1
    },
    {
      "id": 22555,
      "title": "Help to explain cross-validation results",
      "comment_count": 11,
      "views": 1050,
      "votes": 4
    },
    {
      "id": 22651,
      "title": "Weird convergence pattern in convnet (phase transition?)",
      "comment_count": 4,
      "views": 751,
      "votes": null
    },
    {
      "id": 22637,
      "title": "reproducible results with keras/theano",
      "comment_count": 8,
      "views": 1520,
      "votes": 1
    },
    {
      "id": 22581,
      "title": "Some test images are partially obscured",
      "comment_count": 3,
      "views": 659,
      "votes": 5
    },
    {
      "id": 22702,
      "title": "Is the Armadillo C++ library eligible for use here?",
      "comment_count": 1,
      "views": 329,
      "votes": null
    },
    {
      "id": 22732,
      "title": "order of images in submission files",
      "comment_count": 1,
      "views": 302,
      "votes": 1
    },
    {
      "id": 22449,
      "title": "Use dice coefficient as loss function, the cnn seems not learning at all, output all 0s or 1s",
      "comment_count": 5,
      "views": 1676,
      "votes": null
    },
    {
      "id": 22634,
      "title": "dice coefficient",
      "comment_count": 4,
      "views": 740,
      "votes": 1
    },
    {
      "id": 22591,
      "title": "Is Octave 4.0.2 and its standard function libraries eligible for use for this competition?",
      "comment_count": 4,
      "views": 561,
      "votes": null
    },
    {
      "id": 22260,
      "title": "CNN in practice",
      "comment_count": 8,
      "views": 1824,
      "votes": null
    },
    {
      "id": 22588,
      "title": "What are other competitions with Dice coefficient metric?",
      "comment_count": 0,
      "views": 299,
      "votes": 0
    },
    {
      "id": 22571,
      "title": "Looking for team merge",
      "comment_count": 2,
      "views": 329,
      "votes": null
    },
    {
      "id": 22572,
      "title": "Localization",
      "comment_count": 0,
      "views": 184,
      "votes": 0
    },
    {
      "id": 22543,
      "title": "Public and private data subset split. How public data subset(20%) was chosen from all test dataset?",
      "comment_count": 3,
      "views": 512,
      "votes": 3
    },
    {
      "id": 21831,
      "title": "test data is not accurate",
      "comment_count": 41,
      "views": 5656,
      "votes": 6
    },
    {
      "id": 22361,
      "title": "Weighted binary crossentropy",
      "comment_count": 6,
      "views": 1598,
      "votes": 5
    },
    {
      "id": 22424,
      "title": "Is there good way to compensate the class imbalance?",
      "comment_count": 0,
      "views": 424,
      "votes": 0
    },
    {
      "id": 22418,
      "title": "Using data for scientific publications",
      "comment_count": 1,
      "views": 368,
      "votes": null
    },
    {
      "id": 21081,
      "title": "non continuous score function",
      "comment_count": 15,
      "views": 3118,
      "votes": 10
    },
    {
      "id": 22187,
      "title": "question about the dice coefficent",
      "comment_count": 2,
      "views": 437,
      "votes": null
    },
    {
      "id": 22157,
      "title": "What do we need to analyse here",
      "comment_count": 1,
      "views": 532,
      "votes": -1
    },
    {
      "id": 22014,
      "title": "GPU-accelerated Theano & Keras on Windows 10 native",
      "comment_count": 6,
      "views": 5391,
      "votes": 4
    },
    {
      "id": 22062,
      "title": "deformation techniques for image data augmentation",
      "comment_count": 3,
      "views": 1426,
      "votes": null
    },
    {
      "id": 21900,
      "title": "Test-Train Split?",
      "comment_count": 16,
      "views": 1879,
      "votes": 1
    },
    {
      "id": 21093,
      "title": "run-length encoding",
      "comment_count": 2,
      "views": 783,
      "votes": 2
    },
    {
      "id": 21924,
      "title": "Validation score doesn't match LB score",
      "comment_count": 13,
      "views": 1387,
      "votes": null
    },
    {
      "id": 22060,
      "title": "Is scoring a representative sample?",
      "comment_count": 2,
      "views": 645,
      "votes": null
    },
    {
      "id": 22044,
      "title": "naive question?",
      "comment_count": 4,
      "views": 928,
      "votes": null
    },
    {
      "id": 21939,
      "title": "What kind of computer do you use in this kind of competition?",
      "comment_count": 5,
      "views": 1637,
      "votes": null
    },
    {
      "id": 21970,
      "title": "Displaying images during training with a spatial convolutional network",
      "comment_count": 0,
      "views": 471,
      "votes": 1
    },
    {
      "id": 21959,
      "title": "Cannot download train.zip or test.zip",
      "comment_count": 0,
      "views": 347,
      "votes": 0
    },
    {
      "id": 21898,
      "title": "How to perform correctly shift of image?",
      "comment_count": 2,
      "views": 649,
      "votes": null
    },
    {
      "id": 21890,
      "title": "Can the submission file size tell you if you get a good score",
      "comment_count": 1,
      "views": 427,
      "votes": 1
    },
    {
      "id": 21856,
      "title": "Semi-supervised learning and test sync",
      "comment_count": 3,
      "views": 738,
      "votes": 1
    },
    {
      "id": 21495,
      "title": "Is the training data accurate?",
      "comment_count": 7,
      "views": 1972,
      "votes": 5
    },
    {
      "id": 21784,
      "title": "Something I'm thinking",
      "comment_count": 0,
      "views": 546,
      "votes": 2
    },
    {
      "id": 21734,
      "title": "submissions timing out?",
      "comment_count": 0,
      "views": 277,
      "votes": 0
    },
    {
      "id": 21374,
      "title": "Sliding window vs whole image?",
      "comment_count": 3,
      "views": 1406,
      "votes": 2
    },
    {
      "id": 21326,
      "title": "What is train mask CSV good for?",
      "comment_count": 2,
      "views": 989,
      "votes": null
    },
    {
      "id": 21065,
      "title": "Another image processing competition on Kaggle",
      "comment_count": 30,
      "views": 7900,
      "votes": 5
    },
    {
      "id": 21283,
      "title": "Large data files, Python on cloud",
      "comment_count": 3,
      "views": 673,
      "votes": null
    },
    {
      "id": 21185,
      "title": "Non-Commercial Libraries",
      "comment_count": 2,
      "views": 902,
      "votes": null
    },
    {
      "id": 21149,
      "title": "What do you actually predict?",
      "comment_count": 9,
      "views": 2207,
      "votes": 1
    },
    {
      "id": 21220,
      "title": "ImageStat Module to Calculate Statistics in Masked Region of Images",
      "comment_count": 4,
      "views": 1016,
      "votes": null
    },
    {
      "id": 21235,
      "title": "Unknown sponsor?",
      "comment_count": 1,
      "views": 869,
      "votes": null
    },
    {
      "id": 21193,
      "title": "No experience in image processing",
      "comment_count": 3,
      "views": 1336,
      "votes": 1
    },
    {
      "id": 21170,
      "title": "Errors in training image labeling",
      "comment_count": 3,
      "views": 841,
      "votes": 1
    },
    {
      "id": 21176,
      "title": "Data Interpretation",
      "comment_count": 0,
      "views": 561,
      "votes": 0
    },
    {
      "id": 21096,
      "title": "some preliminary results for discussion",
      "comment_count": 3,
      "views": 2161,
      "votes": 13
    },
    {
      "id": 21075,
      "title": "Description of the imaging process and medical info",
      "comment_count": 4,
      "views": 2015,
      "votes": 5
    },
    {
      "id": 21126,
      "title": "Placed in empty submission benchmark",
      "comment_count": 6,
      "views": 791,
      "votes": 1
    },
    {
      "id": 21069,
      "title": "image \"stack\"",
      "comment_count": 3,
      "views": 1131,
      "votes": 1
    },
    {
      "id": 21102,
      "title": "Original RF or IQ data ",
      "comment_count": 0,
      "views": 536,
      "votes": 1
    },
    {
      "id": 21097,
      "title": " Why remove the metadata of the images ?",
      "comment_count": 1,
      "views": 744,
      "votes": null
    },
    {
      "id": 21092,
      "title": "How is the rough recognition?",
      "comment_count": 0,
      "views": 612,
      "votes": 0
    },
    {
      "id": 21076,
      "title": "Additional Rules page?",
      "comment_count": 2,
      "views": 719,
      "votes": null
    },
    {
      "id": 21068,
      "title": "What is \"1 1\" in the submission file?",
      "comment_count": 3,
      "views": 1041,
      "votes": null
    }
  ],
  "errors": []
}