Facebook can now search photos by content

Facebook can now search photos by content

Facebook expanded the possibilities of visual search in social networks using their platform of computer vision Lumos. Now users can search for photos using key words describing their content.

Facebook expanded the possibilities of visual search in social networks using their platform of computer vision Lumos. Now users can search for photos using key words describing their content.

photo search nn diagram

Initially, the purpose of creation of Lumos was to help people with vision problem to understand the content of photos. Subsequently, the company decided to use these developments to all users.

Artificial intelligence system created by Facebook, is able to determine from photo objects, environment, animals, places and clothes. For example, if user a search for [black t-shirt photo]. Lumus will find all photos that contain black t-shirt – even in the absence of labels or descriptions.

"We have created a system that uses image recognized to quickly find the most relevant photos," reads the Facebook post.

The company also noted that the use of machine learning allows us to continually improve the quality of the system.

Google news View all news
URL parameters of the Google search engine results page
There is a possibility to adjust the results for the base query (which corresponds to the value of the parameter q or as_q ) using a number of get-parameters of the format variable=value used in the URL of the page of search results after the substring /search? . The use of these parameters can be very useful when parsing search results.
Read More →
The basic SEO requirements
Here are the list of the most important SEO requirements that must be followed in each optimized page on the site. Requirements are also useful as a SEO guide for the developers of website, which will be optimized.
Read More →
Most popular in Google
Google IR score: what is it and how to use for optimization?