visual image 0.00325524
image content 0.002918031
image corpus 0.002822626
image caption 0.002797022
image captions 0.0026998250000000003
image description 0.002683039
image descriptions 0.002679953
query image 0.002657036
new image 0.002626809
image similarity 0.002603573
image database 0.00259064
first image 0.002504562
example image 0.002479024
original image 0.002463996
global image 0.00246047
second image 0.002427488
noisy image 0.002420131
image descriptors 0.002403534
saliency image 0.002400566
image con 0.00238841
third image 0.002368685
image contents 0.002367971
image retrieved 0.002363214
tiny image 0.002361822
actual image 0.0023601200000000003
visual visual 0.00222072
image 0.00214488
visual information 0.001979217
visual content 0.001883511
visual human 0.001799171
visual score 0.001620271
different images 0.001612255
other images 0.001529823
similar images 0.00147545
images figure 0.001431623
web images 0.001408162
visual differences 0.001392108
visual orig 0.001376715
visual saliency 0.0013660459999999999
visual detectors 0.0013572189999999998
visual google 0.001348583
visual classifiers 0.0013473279999999999
visual estimates 0.0013300229999999998
relevant images 0.0012929299999999999
such data 0.0012689939999999999
online images 0.001255826
similar content 0.001245891
labeled images 0.001244438
word pairs 0.0012330589999999999
retrieved images 0.001221044
images orga 0.001219343
full model 0.0012033629999999998
output sentence 0.001194383
training data 0.001185008
constraints text 0.001183314
human language 0.0011792019999999999
new corpus 0.001159675
information gap 0.001153561
ing corpus 0.001133729
content selection 0.0011187340000000001
sentence compression 0.001113383
visual 0.00111036
level word 0.001109006
caption generation 0.001096655
sentence generalization 0.001093028
text pairs 0.0010867469999999999
circumstantial information 0.001086622
neous information 0.001086622
sentence com 0.001080429
many words 0.001062048
training corpus 0.001059165
caption problem 0.001058475
low content 0.001054406
language descriptions 0.0010254639999999998
automatic caption 0.001023035
benchmark data 0.001019769
age caption 0.001010435
images 0.00100271
text parallel 9.92775E-4
content alignment 9.91049E-4
content misalignment 9.8887E-4
dependency constraints 9.86119E-4
parallel corpus 9.79292E-4
evaluation results 9.77187E-4
original caption 9.71258E-4
descriptive text 9.71137E-4
caption retrieval 9.70283E-4
similarity score 9.686040000000001E-4
gist feature 9.67427E-4
new task 9.598E-4
model 9.51919E-4
future words 9.45492E-4
different methods 9.43458E-4
descriptive caption 9.3205E-4
automatic captions 9.25838E-4
google corpus 9.15969E-4
other work 9.11651E-4
input caption 9.07692E-4
caption generalization 9.01342E-4
semantic matching 9.004810000000001E-4
