Facebook's SEER enhances AI's computer vision
Facebook announced that it had built an artificial intelligence program called SEER that could see what it was searching for, and it did so by feeding it with more than a billion public photos from Instagram.
Instagram has become one of the largest image databases on the planet over the past decade, and the owner of the company, Facebook, is using this treasure trove to teach machines what is in the picture.
Facebook said that the SEER program outperformed current AI models in the Object Recognition Test.
The software scored 84.2 percent on the classification accuracy scale when running a test provided by ImageNet, a large visual database designed for use in visual object recognition software research.
This criterion mainly tests whether the AI software can determine what is in the image.
While many AI models are being trained on categorized datasets, Facebook said SEER learned to recognize objects in images by analyzing random, unclassified, and unsaturated Instagram images.
This AI technique is known as self-supervised learning, Facebook researchers wrote: The future of AI is to create systems that can learn directly from whatever information is presented without relying on coordinated data sets to teach them how to recognize objects in an image or interpret a piece of text.
They added: SEER performance demonstrates that self-supervised learning can excel in real-world computer vision tasks, and this achievement paves the way in the future for more flexible and accurate computer vision models.
While this is only a research project, a Facebook spokesperson said: The potential uses are relatively broad.
The company explained that potential uses include improved auto-generated text to describe images for people with visual impairments, better automatic classification of items sold through the Facebook marketplace, and better systems to keep harmful images away from the Facebook platform.
Priya Goyal, a software engineer at Facebook's Artificial Intelligence Research Lab, said: We are informing Instagram account holders via a data policy that we are using the information we have to support research and innovation.
"Although we are sharing the details of our research and building an open source library that allows other researchers to use self-learning to train models on images, we do not share images or create SEER," Goyal added.
Other big tech companies, including Google and Microsoft, are also trying to push the boundaries of computer vision.
And last summer, Google released the SimCLRv2 computer vision model, while OpenAI published the iGPT 2.