In Barcelona, Spain, an invitation-only event hosted a variety of different professionals working in artificial intelligence.
Apple was part of that event, including the company’s head of machine learning division, Russ Salakhutdinov, and other employees. The talk was focused on machine learning, and progress made in AI. Thanks to Quartz, the talk and slides have been revealed.
There are a variety of different topics covered in the process, including volumetric detection of LiDAR:
“One presentation slide that summarized the company’s research featured two pictures of cars, to illustrate “volumetric detection of LiDAR” and “prediction of structured outputs.” Both LiDAR, or Light Detection and Ranging (similar to radar but with lasers), and prediction of physical events are important building blocks of today’s self-driving car technologies. However, two attendees of the presentation, who asked not to be identified due to the sensitive nature of the content, both stressed that the company made no mention of cars or automotive ambitions.”
On top of that, Apple also talked about image processing and image recognition algorithms, which are now capable of processing 3,000 images per second.
One other major topic is neural networks, where Apple is working on shrinking them down so that they can work directly on devices moving forward:
“Another slide focused on Apple’s ability to build neural networks that are 4.5 times smaller than the originals with no loss in accuracy, and twice the speed. The technique, not unknown in AI research, uses a larger, more robust neural network to teach another network the decisions it would make in a variety of situations. The “student” network then has a streamlined version of the “teacher” network’s knowledge. In essence, it predicts the larger network’s predictions about a given photo or audio sample.”
Apple is going to open its gates when it comes to AI and its researchers as well, allowing them to publish their work for peer review. This should not only allow Apple to improve its use of AI, but also not restrict potential new hires who were previously worried about not being able to publish their work.