The famous social media platform, Facebook has announced on Thursday that the company has created an artificial intelligence (AI) program that can recognize images on its own without any provided labeling. The program was able to achieve this by feeding it over 1 billion unrestricted pictures from Instagram.
The “computer vision” program, named as “SEER”, outperformed current AI models in an article identification test, commented Facebook.
As per the reports, the program has accomplished a “classification accuracy score” of 84.2 percent, when the program was examined by ImageNet, which is a large visual database created for application in visual object recognition software research. In general, the database tests and identifies AI programs for their features.
“The future of AI is in creating systems that can learn directly from whatever information they’re given — whether it’s text, images, or another type of data — without relying on carefully curated and labeled data sets to teach them how to recognize objects in a photo, interpret a block of text, or perform any of the countless other tasks that we ask it to,” Facebook’s researchers wrote in a blog post.
“SEER’s performance demonstrates that self-supervised learning can excel at computer vision tasks in real-world settings,” they continued. “This is a breakthrough that ultimately clears the path for more flexible, accurate, and adaptable computer vision models in the future.”
“We inform Instagram account holders in our data policy that we use the information we have to support research and innovation including in technological advancement like this,” stated Priya Goyal, a software engineer at Facebook AI Research.
Moreover, other renowned tech companies including Google and Microsoft are in the queue for pushing the boundaries of computer vision. Previously, Google has published the SimCLRv2 computer vision model, whereas OpenAI has published iGPT 2.