Last month, Pinterest unveiled their latest discovery tool called Pinterest
Lens, which enables users to take photos of real world objects and find matches
for them on Pinterest based on image recognition technology. To access
Pinterest Lens, you first need to update your Pinterest app to the latest
version. Once you’ve done that, you can access Lens by tapping on the search
bar, then tapping the red camera icon in the top right of screen. This will open
the lens camera – point it at whatever object you want to search for, wait a
moment for the system to match the image, and you’ll be presented with related
Pin options. Pinterest notes that, as they had anticipated, initial tests have
shown that most users are using Lens to identify objects relating to fashion,
home décor, arts and food.
Along this line, Pinterest is also giving users a way to contribute to the
development of Lens, with an option to manually tag objects they identify. This
will help Pinterest evolve Lens faster, by essentially crowd-sourcing object
identification, giving them a larger pool of tagged sample images to work from.
Adding to the search functionality of Lens, Pinterest is also giving users
another way to search for related objects, with a new section within Lens which
presents ideas to try out and see what matches show up. This will
highlight new objects as Pinterest adds them in, which will also give the
Pinterest engineering team another way to boost their understanding of what
people are using Lens for by analyzing what matches people most commonly
click on from these examples.
Lens is an interesting addition for Pinterest, and one which should get
them more attention. If users get good results, maybe links to relevant Pins
that can help them save money on products by identifying similar items, the
innovation could help Pinterest add to their active user count and continue
their expansion. Image recognition tools are set to become a much bigger
consideration in digital marketing, with Snapchat, Twitter, Facebook and Google all
working to improve their image recognition tools and options. Soon, the ability
to search images will be commonplace.
No comments:
Post a Comment