Human Generated Data

Title

Levantine textile

Date

People
Classification

Textile Arts

Human Generated Data

Title

Levantine textile

People
Date

Classification

Textile Arts

Machine Generated Data

Tags

Amazon

Embroidery 98.2
Pattern 98.2
Rug 96.2
Stitch 74.9
Art 58.1
Applique 57.6

Clarifai

decoration 98.8
pattern 96.8
design 96.5
art 95.7
retro 92.6
vintage 92.2
flower 91.2
texture 90.6
desktop 89.8
ornate 89.1
travel 87.9
style 86.4
illustration 86.3
culture 86.1
wall 84.2
traditional 84.2
old 84
antique 83.7
abstract 82.5
symbol 82.4

Imagga

flag 100
emblem 75
paper 29.8
retro 27.9
old 27.2
money 24.7
currency 23.3
vintage 22.3
grunge 20.4
texture 20.1
finance 19.4
cash 19.2
design 18.6
art 18.2
antique 18.2
letter 17.4
decoration 17.3
ancient 17.3
symbol 16.8
note 16.5
envelope 16.5
pattern 16.4
bank 16.1
stamp 15.5
national 15.4
wallpaper 15.3
pillow 14.8
banking 14.7
graphic 14.6
fabric 14.5
bill 14.3
card 13.8
container 13.7
mail 13.4
frame 13.3
aged 12.7
pay 12.5
country 12.3
cushion 12.2
postage 11.8
dirty 11.7
announcement 11.6
patriotic 11.5
sign 11.3
color 11.1
business 10.9
postmark 10.8
collection 10.8
wealth 10.8
financial 10.7
banknote 10.7
patriotism 10.7
notes 10.5
clip 10.2
backdrop 9.9
postal 9.8
news 9.7
payment 9.6
state 9.6
wall 9.5
nation 9.5
culture 9.4
economy 9.3
decorative 9.2
material 8.9
newsflash 8.9
update 8.9
latest 8.9
breaking 8.9
announce 8.9
rippling 8.8
waving 8.8
textured 8.8
wavy 8.7
splash 8.6
international 8.6
ripple 8.6
post 8.6
padding 8.4
savings 8.4
top 8.3
element 8.3
border 8.1
history 8.1
information 8
stamps 7.9
holiday 7.9
colorful 7.9
banknotes 7.8
drawing 7.7
used 7.7
parchment 7.7
exchange 7.6
sheet 7.5
rich 7.4
page 7.4
graffito 7.4
map 7.3
paint 7.2
celebration 7.2
decor 7.1

Google

Textile 74.7
Window valance 73.5
Font 68.6
Banner 55.7
Art 50.2

Microsoft

painting 94.2
art 72.7

Feature analysis

Amazon

Rug 96.2%

Captions

Microsoft

a close up of a rug 60.1%
a close up of a curtain 38.4%
close up of a rug 38.3%

Text analysis

Amazon

mmemwe