Human Generated Data

Title

[Train and utility lines]

Date

1936-1937

People

Artist: Lyonel Feininger, American 1871 - 1956

Classification

Photographs

Credit Line

Harvard Art Museums/Busch-Reisinger Museum, Gift of T. Lux Feininger, BRLF.351.12

Copyright

© Artists Rights Society (ARS), New York / VG Bild-Kunst, Bonn

Human Generated Data

Title

[Train and utility lines]

People

Artist: Lyonel Feininger, American 1871 - 1956

Date

1936-1937

Classification

Photographs

Machine Generated Data

Tags

Amazon
created on 2019-05-29

Pollution 95.1
Smoke 92.1
Vehicle 90
Transportation 90
Train 90
Railway 79.1
Train Track 79.1
Rail 79.1
Nature 77.9
Building 65.9
Human 64.7
Person 64.7
Locomotive 60.1
Fog 59

Clarifai
created on 2019-05-29

vehicle 99.6
transportation system 99.3
people 98.1
watercraft 95.6
military 94.8
aircraft 93.5
railway 92.8
train 92.6
many 92.4
no person 92.1
group 91.6
group together 90.6
war 90.5
calamity 84.1
airport 84
monochrome 80.9
administration 79.4
smoke 79
street 78.9
industry 78.8

Imagga
created on 2019-05-29

freight car 100
car 94
wheeled vehicle 73.5
vehicle 53.1
conveyance 24.9
transportation 19.7
travel 19
vessel 18.7
building 17
wall 16.6
architecture 14.8
old 14.6
road 13.6
snow 13.3
urban 13.1
industry 12.8
sky 12.8
landscape 12.6
track 12.5
water 12
street 12
winter 11.9
industrial 11.8
train 11.7
city 11.6
station 11.6
tub 11.2
cold 11.2
transport 11
ship 10.3
light 10
environment 9.9
structure 9.2
metal 8.9
steel 8.8
rail 8.8
railway 8.8
season 8.6
grunge 8.5
black 8.4
shipping 8.3
dirty 8.1
river 8
night 8
submarine 7.9
scene 7.8
weather 7.7
line 7.7
construction 7.7
concrete 7.7
traffic 7.6
dark 7.5
lights 7.4
window 7.3
lake 7.3
business 7.3
room 7.2
trees 7.1
sea 7
modern 7

Google
created on 2019-05-29

Microsoft
created on 2019-05-29

train 99.4
outdoor 98.4
fog 93.9
vehicle 92.8
land vehicle 90
locomotive 83.2
sky 79.7
white 75.5
steam 70.2
black and white 58
old 48.2
railroad 27.5

Feature analysis

Amazon

Train 90%
Person 64.7%

Captions

Microsoft

a black and white photo of a train 86.1%
black and white photo of a train 80.9%
an old photo of a train 80.8%