Машинне навчання та доповнена реальність на пристроях на базі iOS із фреймворком MLARKit

dc.contributor.authorГороховський, Семен
dc.contributor.authorФранків, Олександр
dc.description.abstractСтворено фреймворк MLARKit, що дає змогу легко користуватися складними для використання у мобільних пристроях алгоритмами машинного навчання та доповненої реальності, враховуючи їхні особливості. Фреймворк створено максимально гнучким, тож сторонні розробники зможуть максимально задовольнити свої потреби без самостійної реалізації.uk_UA
dc.description.abstractMachine learning is a technology that requires a significant amount of computing power and therefore is not efficient in terms of energy consumption. The equivalent may be applied to the technology of augmented reality which impacts in contrast not only the processor but also the camera, accelerometer, and gyroscope. Moreover, due to the work with the visual effects, the speed of computing truly matters to the latter. In recent years it has become possible to use these technologies on mobile devices due to the grand efficiency optimizations. The problem these days is that it requires an imposing amount of resources for engineers to spend on the development of the experiences with the use of technologies mentioned above. The solution developed is the next step in simplifying the process of integration of such experiences in mobile applications for the iOS operating system. While Core ML and AR Kit provide a simple and comprehensive interface for the basic machine learning and augmented reality routines, the new framework – Machine Learning Augmented Reality Kit (MLARKit) – provides the same simplicity and flexibility for the engineers in terms of the two tasks: replacing real objects with the virtual ones and marking of recognized objects or images in augmented reality which are widely used in modern mobile applications as the additional functionality. With the use of the above-mentioned framework, the integration of such experience comes down to the conformance of the only protocol which provides a full range of settings. This article covers the problems that may be faced while solving power-consuming and energy inefficient tasks on mobile devices as well as the approaches used to reduce such impact. The manner in which the newly developed framework is designed is described in the article as well. As an improvement, it is important to consider tracking the moving objects which may reduce the amount of computing power consumed significantly.en_US
dc.identifier.citationГороховський С. С. Машинне навчання та доповнена реальність на пристроях на базі iOS із фреймворком MLARKit / Гороховський С. С., Франків О. О. // Наукові записки НаУКМА. Комп'ютерні науки. - 2020. - Т. 3. - С. 4-6.uk_UA
dc.relation.sourceНаукові записки НаУКМА. Комп'ютерні науки.uk_UA
dc.statusfirst publisheduk_UA
dc.subjectмашинне навчанняuk_UA
dc.subjectCore MLuk_UA
dc.subjectдоповнена реальністьuk_UA
dc.subjectAR Kituk_UA
dc.subjectмобільні пристроїuk_UA
dc.subjectmachine learningen_US
dc.subjectCore MLen_US
dc.subjectaugmented realityen_US
dc.subjectAR Kiten_US
dc.subjectmobile devicesen_US
dc.titleМашинне навчання та доповнена реальність на пристроях на базі iOS із фреймворком MLARKituk_UA
dc.title.alternativeMachine Learning and Augmented Reality on iOS Devices with MLARKit Frameworken_US
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
387.52 KB
Adobe Portable Document Format
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
7.54 KB
Item-specific license agreed upon to submission