Visual objects
identification and attribution
solution

Loading

Richard
O’Connor

Entrepreneur

Richard is a 35 year old entrepreneur living in a big city.

Richard has a strong personality and value set, and
considers his style very defined. He knows the style and
type of outfits that suit his taste; he is always frustrated
by the lack of relevant personalization and off-the-mark
similar looks suggestions while shopping online.

Richard knows what he wants, and doesn't want to spend
his time searching for the right things. He values
convenient, effortless and time efficient shopping.

ricgard

Similar looks

  • 3x faster shopping
  • Effortless search
    of favored products
  • Personalized
    recommendations
  • Ability to purchase “full look”
    of preferred style

Suzan Smith

Fashion Blogger

Suzan is 27 years old, she is a popular fashion
blogger on Instagram.

Suzan is a top fashion expert and a heavy shopper,
and she is a fashion “icon” for her followers.
She considers looking stylish every day to be her job; it
is of key importance to her success!


Suzan follows all popular and emerging trends; she
needs to have an ability to shop for the looks she
finds online to keep up with the fashion.

ricgard

Visual Search

  • Quick accurate
    visual search
  • Ability to pick among the
    variety of products
    of similar style
  • Instant purchase
    of wanted outfits

Steve Collins

Head of Digital Marketing
Apparel


Steve is 52 years old, he is an executive in a large apparel
brand.

As the Head of Digital Marketing, Steve is always looking
for the ways to boost the business and grow revenue, as
well as key commerce metrics and customer loyalty.


In today's digital world, Steve understands they have to
heavily employ technology to stay competitive in the
market.

ricgard

Next gen search

and recommendations
engine adopted by Steve

UP TO

15%

increased Return
Customer rate

200%

ROI within 3 months

Significant

conversion rate
increase

PLUS

Increased AOV
Technological advantage & innovation among competitors

Automated

SKUs management

  • Automated
    products attribution
  • Savings of up to 40 people
    hours per month
  • Positive impact on employees
    satisfaction by removing
    routine work

TRY IT YOURSELF

RESULTS
Tap to open Camera Roll

TRY IMAGES FROM OUR GALLERY

slider image slider image slider image slider image slider image slider image slider image slider image slider image slider image slider image

visual objects
identification and attribution
solution

Data Science solution that can:

Identify

any object in
images/photos

Categorize

objects and
determine their
attributes

Find

visually similar
objects in existing
data sets

Automate

content
management and
moderation

how it
works

  • Deep neural networks and fine-tuning
    were used to train models
  • Recommendations are performed
    based on deep representations
    of images
  • A couple hundred images per class
    are enough for 95% accuracy of
    models
  • Models could be easily run as micro
    services
    and integrated with any
    existing systems
  • Implemented using
    Python/TensorFlow/GPU computing