Microsoft appears to be considering integrating emotional and cognition-based tools into professional recording equipment using Hololens, based on IP documentation reviewed by AndroidHeadlines. Filed under categories for studio equipment and devices that control 'television cameras', the patent doesn't currently contain any imagery or in-depth documentation to clearly define the invention. Based on its description, Microsoft is looking to utilize sensor technology similar to what might typically be found in its Hololens AR headset in order to gauge the emotional or cognitive state of the wearer or a subject. A 'buffer' temporarily stores video content that is captured based on related triggers and holds onto that until a change in the emotional or cognitive state is detected by the system. That, in turn, sets off the creation of a video segment based on the stored footage.
Background: Microsoft's solution here would most likely be based in some form of AI and machine vision but it wouldn't be the first company to explore automation of tasks based on emotional or cognitive states. Huawei revealed that it was working on its own AI capable of accurately tracking human emotion back in April. That's being undertaken with the goal of making AI more comprehensive since it may be able to better assist users if it can understand them beyond simple voice commands. Although the impact and use cases for that particular iteration of the technology aren't immediately clear, it is well known that human communication is at least as dependent on reading facial expression and body language as spoken language. So an AI that can read emotions could feasibly be taught to be more nuanced and more helpful.
Samsung is another company that's been considering the implications of an AI that reads human cognition but with concepts related to shopping. Its system would aspire to understand a user's emotional response to shopping items in order to offer up better recommendations. That's as opposed to depending on a user's words, which might not provide enough details comparatively to account for nuances there. While a user can readily say that they 'like' or 'dislike' something, that doesn't really say 'how much' they like it or dislike it. Even systems that allow for tiered responses are relatively rigid with regard to exactly how much a user loves or hates something.
Impact: The possible uses for and complexity of Microsoft's own invention on the emotion-tracking front appear more straightforward on the surface than either abovementioned example of emotion or cognition-based tracking technology. That may not actually be the case though. The company is obviously considering a tool that will record and playback video footage based on the emotional state of the subject but it isn't really apparent how that might be used. The described ability to generate footage showing different states of emotion each time that changes could have practical applications in the editing process. In filming, scenes often have a tone that stays relatively constant throughout. Rather than depending on hardware controls to capture a scene, the technology could automatically separate scenes based on the emotion being presented. Perfecting that, depending on the complexity of the scene in question could prove extremely troublesome but that's just one way the new tech might be used if Microsoft ever puts its AR-based patent to use.