Using "pornography or advertisement" as an example
{
    // The outcome of task "porn", the "54bcfc6c329af61034f7c2fc" is task id.
    '54bcfc6c329af61034f7c2fc': {
        statistic: [0, 0, 1], // The amount of images of "porn","sexy","normal" respectively
        reviewCount: 0, // The amount of images needs to be review.
        // fileList:fileList: outcome of each image.
        fileList: [{
            rate: 0.9927366971969604, // A float number between 0-1, represent the possibility of a specific label. higher the better.
            label: 2, // The outcome label, here "2" means "normal"
            name: 'http://ztest.qiniudn.com/gogopher.jpg', // the `name` input.
            tag: 'some_tag', // 【Optional】the `tag` input
            zip: 'zip_file_name', // 【Optional】if the `image` is a archive, every file in it will get the archive name.
            review: false // Need review or not
        }]
    },
    // The outcome of task "advertisement", the "5588dba4c7ee53a04b5fad7d" is the task id.
    '5588dba4c7ee53a04b5fad7d': {
        statistic: [1, 0, 0, 0],
        reviewCount: 0,
        fileList: [{
            rate: 0.9999947547912598,
            label: 0,
            name: 'http://ztest.qiniudn.com/gogopher.jpg',
            tag: 'some_tag',
            zip: 'zip_file_name',
            review: false
        }]
    },
    nonce: '0.010413094889372587',
    timestamp: 1435385737,
    code: 0,
    message: 'success'
}
    
Descriptions:
Name Type Descriptions
fileList Array

[returned on success] An array of objects, an example object instance: {rate: xxx, label: 2, review: false, name: ”00003.jpg”}

rate: A floating point number ranged from 0 to 1. This is the probability that the image is classified into a class.

label: A non-negative integer representing the class that the image has been classified into. See Interfaces for more details.

Note that when label is -1, it means there are some errors in the image or its format is not supported.

review: A Boolean value indicating whether manual review is needed for the image. When the computer is more than 99.5% certain of the result, and no manual review is needed, its value is false, otherwise it is set to true. The threshold for manual review is configurable.

name: The raw filename. For clarity, this value should be unique in the same batch. We suggest using unique identifiers for all images.

statistic Array [returned on success] The array that stores the number of images classified into each class.
reviewCount Number [returned on success] The total number of images that might need manual review.

Using “Pornography recognition + violence recognition analysis” as an example
{
    // The outcome of task "porn", the "54bcfc6c329af61034f7c2fc" is task id.
    '54bcfc6c329af61034f7c2fc': {
        // Outcome of a bundle of images,As the example:50.0% might be normal; 25.53% might be porn; 24.46% might be sexy
        avgScores: [
            {score: 0.5000619335640717, label: 2},   // "label:2 -> Normal"
            {score: 0.25529559658062695, label: 0},  // "label:0 -> Porn"
            {score: 0.24464247275246542, label: 1}], // "label:1 -> Sexy"

        // The extremum represent the maximum score of each label, the absence of the the value means no such label.
        // The main purpose of extremum value is to provide a means to monitor the outliner you might care about.
        // for example if you care about "porn" very much, you should check the score of label-0 in extremum values.
        maximums: [
            {max: 0.9999999220340654, label: 0},  // in this case, the maximum value of the label-0 in extremum sets is 0.999999922, that mean we are very positive there is porn, watch out!
            {max: 0.9999772906303406, label: 2},  // as above, the normal extremum is also very high(0.9999772906303406), but we don`t care
            {max: 0.9784060716629028, label: 1}], // as above again, the sexy is 0.9784060716629028

        // The amount of image of each label
        statistics: [
            {count: 2, label: 2},  // 2 normals
            {count: 1, label: 1},  // 1 sexy
            {count: 1, label: 0}], // 1 porn

        // The entire video is labeled "porn" because ther is 1 porn appeared
        label: 0,
        // If the outcome should be reivew or not.
        review: false,
        topN: 3
    },

    // The outcome of task "terrorism", the "55c03cf238dc1cfb3d80be14" is task id
    // label-0: normal; 1: terrorism
    '55c03cf238dc1cfb3d80be14': {
        avgScores: [
            {score: 0.874534472823143, label: 0},
            {score: 0.1254655198135879, label: 1}],
        maximums: [
            {max: 0.9992490410804749, label: 0},
            {max: 0, label: 1}],
        statistics: [
            {count: 4, label: 0},
            {count: 0, label: 1}],
        label: 0,
        review: false,
        topN: 2
    },
    nonce: '0.809460110263899',
    timestamp: 1442636213,
    code: 0,
    message: 'success'
}
    

Examples of object detection and recognition tasks return results.
    {
        //The outcome of "object Identification" task, "5673875659a518eb6c2af055" is the task id
        '5673875659a518eb6c2af055':{
            fileList:[{
                name:  'http://img4.imgtn.bdimg.com/it/u=1404723066,982370128&fm=206&gp=0.jpg',  //Url or filename
                objects:[
                {    //Coordinate sets,value is proportion to image size
                    location: [     //Coordinates [topleft,topright,bottomright,bottomleft]
                         [ 0.7007799905443949, 0.26091899871826174 ],
                         [ 0.8222528366815476, 0.26091899871826174 ],
                         [ 0.8222528366815476, 0.5064324118874289 ],
                         [ 0.7007799905443949, 0.5064324118874289 ] ],
                    category: "face", // Object type
                    scores: 0.9999884366989136 // score, range【0~1】 the higher the score the better,
                {
                    location: [
                         [ 0.2679353502061632, 0.11118876717307351 ],
                         [ 0.37795620872860863, 0.11118876717307351 ],
                         [ 0.37795620872860863, 0.3357514121315696 ],
                         [ 0.2679353502061632, 0.3357514121315696 ] ],
                    category: "face",
                    scores: 0.9999574422836304 }]
                    }]
        }
        nonce: '0.25289168604649603',
        timestamp: 1431657207,
        code: 0,
        message: 'success'
    }
    
Descriptions:
Name Type Descriptions
fileList   Array

[Returned on success] An array of objects, an example object instance: { name: 'http://img4.imgtn.bdimg.com/it/u=1404723066,982370128&fm=206&gp=0.jpg', objects: [ { location: [ [ 40.825688073394495, 20.333333333333332 ], [ 58.25688073394495, 20.333333333333332 ], [ 58.25688073394495, 33 ], [ 40.825688073394495, 33 ] ] , category: "face", scores: 0.9999574422836304} ] }

The URL or filename of the image from the POST request.

A collection of coordinates of identified object’s key locations.

location: Coordinates of the four corners of the object’s bounded box, whose values are the ratios of the image.

category: Item category.

scores: A floating point number ranged from 0 to 1, denoting the confidence of the recognition. The higher this number is, the more accurate the regnition will be.