Expanding on its efforts to engage with the academic and research communities, Apple Inc. has launched the “Apple Machine Learning Journal,” a blog dedicated to publishing scholarly research papers on the growing field of teaching computers to learn on their own.

“Here, you can read posts written by Apple engineers about their work using machine learning technologies,” a welcoming statement on the site notes, before adding that it wants feedback from researchers, students and developers on the articles published.

The first post on the site, and the only one so far, discusses how to improve realism in synthetic images, or how a computer-generated image can look more like a real image. While it may sound a bit unexciting, the application of machine learning and artificial intelligence to create an image can also be applied in reverse – that is, the same algorithms can be used to identify what is in an existing image as well.

Apple renounced its notoriously tight secrecy on research when it announced in December that it would start publishing its AI and machine learning research. While it would be nice to think that Apple may have decided to share its research out of some benevolent care for the greater good, the decision was nothing of the sort.

Instead, the iPhone maker made the decision to attract more talent. Top researchers routinely like to publish papers and contribute to open-source projects, but Apple had always restricted its employees from doing so, subsequently making it difficult for the company to attract top talent. And as machine learning has become a key differentiator for a raft of new online services, Apple can’t afford to fall behind rivals such as Google Inc., Facebook Inc. and Microsoft Corp.

The company’s first published paper was spotted just after Christmas and perhaps not coincidentally also dealt with machine learning and simulated images. Whether Apple has an obsession with machine learning and imagery isn’t clear. It may well be developing the technology for something like identification scanning on iPhones, or it may simply be the case that it’s not comfortable yet in publishing papers that deal with other facets of its AI development programs.

Photo: Pixabay