11 | 12 |
17 | 18 |
We will not allow each worker to complete and submit more than one of our HITs.
21 | 22 |23 | 24 |
- This HIT contains 10 pages of image selection tasks.
27 | 28 |- Each page contains a task for 1 object class name and 48 candidate images. 29 |
30 | 31 |- Select all images corresponding to the shown object class name.
32 | 33 |- If unsure, you can use the Wikipedia link to learn about the object class.
34 | 35 |- When you are done with a page, click submit and move on to the next page.
36 | 37 |38 | 39 |
- Selected images will appear with blue borders.
45 | 46 |- If necessary, click on the image again to undo the selection.
47 |
48 |
- Do not select drawings or paintings.
55 |
- Red pointer indicates the current position of the mouse cursor.
63 | 64 |- Red pointer will be fixed at the position of the image where you click.
65 | 66 |- When you click on the image again to de-select it, the red pointer will also be released.
67 | 68 |
- Try to click on the object of interest.
74 | 75 |- When there are multiple objects, click on one of them.
79 | 80 |81 | 82 |
- The annotations are performed on a separate web page linked below.
85 | 86 |- Click on the link to open the page and complete 10 pages of tasks.
87 | 88 |- DO NOT CLOSE the current page during the annotation.
89 | 90 |- When you complete the tasks, you will get a unique survey code 91 | (e.g. A2F4C9).
92 | 93 |- Put the survey code in the form at the bottom of this page to claim the reward.
94 | 95 |- You may be blocked from this task if the quality of the submission does not meet the minimal 96 | standard.
97 | 98 |99 | 100 |
Some of the images may include nudity and other disturbing 103 | contents (e.g. snakes, insects, and birds).
104 | 105 |106 | 107 |
108 | 109 | 110 |
116 | 117 |
118 | 119 | 122 | 123 |
34 | You’ve submitted all annotations.
35 |
36 | Put this survey code in AWS Mturk survey form.
37 |
54 | Please drag and drop icons from the bottom panel to matching objects in 55 | the image. If and icon matches multiple objects you can drag the icon 56 | onto any of the objects. There area 11 sets of objects to drag onto the 57 | image. Use the buttons or arrow keys to cycle through them. There are 58 | total of {app.totalSubmitCount} images to label. 59 |
60 | 61 |
73 | {JSON.stringify(imageBoard, null, 2)}
74 |
75 | >
76 | )}
77 |
78 |