brand-detector
Brand detector is a particular case of image classification, since these may contain only text, images, or a combination of both.
In this work, we trained a system for the brand classification of shoes available at https://www.shooos.com/.
The method allows obtaining the most similar brands on the basis of their shape, color, business sector, semantics, general characteristics, or a combination of other features.
The proposed approach is evaluated using 15% of unseen images of the dataset. The experimentation carried out attained reliable performance results, both quantitatively and qualitatively.
How to test:
- find on Google Images any image related to shoe brands below and test it!
This model is not fully trained yet :D
Example Images
adidas
arkk-copenhagen
asics
birkenstock
bjorn-borg
by-garment-makers
camper
carhartt-wip
caterpillar
champion
chpo
clae
colorful-standard
converse
dc-shoes
dedicated
dickies
doughnut
dr-martens
fila
fjallraven
hanwag
happy-socks
havaianas
herschel-supply
iriedaily
keepcup
lefrik
loqi
makia
maloja
new-balance
new-era
nike
norba-clothing
on-running
onitsuka-tiger
palladium
rains
reebok
saucony
secrid
sneaky
stance
superfeet
the-north-face
timberland
toms
triwa
under-armour
vans
veja
- Downloads last month
- 11
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Evaluation results
- Accuracyself-reported0.843