Searching for photos has never been this easy as Facebook latest update has been geared towards this with the aid of its computer vision platform. Searching of photos is done using any keyword that explains the image’s content.
Facebook new search idea is also based on matching your search to posts from your friends that perfectly fits the description. For instance, if you search for “pictures of oranges” or “pictures of flower” the search will yield posts from your friends. The results will be shown higher in the grid accompanied alongside other vital images that you may find applicable.
Taking a closer look at IOS 10 or google photos image-recognition features, the Facebook search update is of similar principle.In this new search feature, abstract keywords also work. I ran a photo search using the keyword “creepy” and these results popped up.
Results are tied with other context clues from images such as caption or comments when it comes to a more subjective search.When I ran a search on “bad photos” images from friends at the women’s march a few weekends ago popped up which showed signs pointing to “bad guys”. Another result showed an apparent fine photo which was captioned describing the subjects as “un photogenic”.
The search wordings are not completely fixed in with graph search which brings results from specific posts from its users. Therefore, the search on “food photos” will not bring same result as a search on “photos of me with food” only if you are tagged in various food selfselfies
The company’s zeal to use computer vision in more consumer facing products is placed in line with the photo search update which is readily available to both web and mobile users in the US.
The previous year the artificial intelligence was launched to aid visually impaired users identify various elements of an image. The A.I will later be applied to analyze context for video searches as this was made public by TechCrunch