Facebook is using Instagram photos to train its image recognition AI

freestocks-org-484218-unsplash.jpg

At its F8 developers conference Facebook not only revealed a number of new Instagram features, the company also talked about how it is using the billions of images on Instagram to train the world's most accurate image recognition systems.

Training deep learning models for image and object recognition is typically a very labor-intensive task, as each training image has to be looked at and labeled by human workers. This is a serious limitation to the size of training image databases; however, Facebook has found a way to reduce human supervision in the training process by using images that are already labeled... with Instagram hashtags.

Its researchers used 3.5 billion Instagram images with approximately 17,000 hashtags to train deep learning models and the results have been encouraging.

A computer vision system that had been trained with one billion images and 1,500 hashtags, achieved 85.4 percent accuracy on the ImageNet benchmarking tool, outperforming the previous leading system by 2.3 percent.

It will be important to manage the disadvantages of less curated labels but the Facebook research shows that less supervised training of image recognition systems could be a step into the right direction, allowing for the use of much larger and databases and therefore improved image and object recognition and classification.

Translation: finding that photo you never tagged that's buried miles deep in your archive might soon get a whole lot easier.



All the photos are from the web and the copyright retains with the original author. If there is any problem, please contact us.
You Might Also Like
Google lens can now identify pet breeds, create pet photo books and compile pet moviesGoogle lens can now identify pet breeds, create pet photo books and compile pet moviesGoogle first introduced Google Lens with the Pixel 2 smartphone. The feature uses Artificial Intelligence to power its visual recognition algorithms and provide information about whatever your smartphone's camera is pointed at, and recently it was made available on iOS devices and integrated in the Google Photos app.​
Google Research explains the Pixel 2 portrait modeGoogle Research explains the Pixel 2 portrait modeBackground-blurring portrait or bokeh modes have pretty much become a standard feature on dual-camera equipped phones. Similar effects can be achieved with single-lens devices but operation tends to be more cumbersome, with more manual interference required, and results less realistic than on dual-camera setups.
Instagram’s new Direct replies are basically threaded messaging for photosInstagram’s new Direct replies are basically threaded messaging for photosInstagram’s Direct messages are becoming a bit more powerful. Following the addition of threaded conversations earlier this week, you can now directly reply to specific individual photos or videos shared with you with your own media.
Apple joins Instagram to show off iPhone photos (of course)Apple joins Instagram to show off iPhone photos (of course)Instagram has 700 million monthly active users, hundreds of millions of which undoubtedly post from their iPhones. Now Apple has joined them, and you'll never guess what phone took its first photos.

Comments

   
Post
more>>