NUS-WIDE-SCENE
The Descriptions of the Files Included in the NUS-WIDE-SCENE
The NUS-WIDE-SCENE is a real world object image dataset. As a subset of NUS-WIDE, it contains 33 scene categories and 34,926 images in total. We use half of the total number (i.e., 17,463 images) for training and the rest for testing. The figure below illustrates the list of scene categories and the numbers of training and testing images.
Total unique number of images in database_object: 34,926
total training image: 17,463 total testing image: 17,463
category
airport - train 88 - test 81
category beach - train 1,438 - test 1,466
category
bridge - train 643 - test 663
category buildings - train 1,991 - test 2,016
category castle - train 75 - test 66
category cityscape - train 481 - test 445
category clouds - train 9,537 - test 9,598
category frost - train 200 - test 200
category
garden - train 373 - test 369
category glacier - train 181 - test 162
category grass - train 2,155 - test 2,209
category harbor - train 149 - test 136
category house - train 561 - test 667
category lake - train 3,448 - test 3,532
category moon - train 245 - test 240
category mountain - train 1,307 - test 1,377
category nighttime - train 505 - test 500
category ocean - train 2,583 - test 2,660
category plants - train 1,398 - test 1,418
category
railroad - train 171 - test 166
category rainbow - train 88 - test 113
category reflection - train 2,081 - test 2,084
category road - train 1,039 - test 1,092
category sky - train 11,995 - test 12,058
category snow - train 1,189 - test 1,132
category street - train 321 - test 319
category sunset - train 2,692 - test 2,720
category temple - train 191 - test 188
category town - train 301 - test 297
category valley - train 817 - test 824
category water - train 7,408 - test 7,465
category waterfall - train 152 - test 153
category window - train 1,212 - test 1,230
-------------------------
raw data
-----------------------------
NUS-WIDE-SCENE
concept list
|
airport |
|
beach |
|
bridge |
|
buildings |
|
castle |
|
cityscape |
|
clouds |
|
frost |
|
garden |
|
glacier |
|
grass |
|
harbor |
|
house |
|
lake |
|
moon |
|
mountain |
|
nighttime |
|
ocean |
|
plants |
|
railroad |
|
rainbow |
|
reflection |
|
road |
|
sky |
|
snow |
|
street |
|
sunset |
|
temple |
|
town |
|
valley |
|
water |
|
waterfall |
|
window |
If
you have any questions about the NUS-WIDE dataset, please contact Dr. Jinhui Tang: