Repozytorium publikacji - Politechnika Gdańska

Today’s smartphones have more and more features that so far were only assigned to personal computers. Every year these devices are composed of better and more efficient components. Everything indicates that modern smartphones are replacing ordinary computers in various activities. High computing power is required for tasks such as image processing, speech recognition and object detection. This paper analyses the performance of the OpenCL (Open Compute Language) environment on mobile devices, which is a library dedicated to high-speed parallel computing. This paper examines how smartphones can access a library that, as it turned out, is not officially supported on the Android platform, and briefly describes the evaluated library. As a part of the study, this API (Application Programming Interface) was tested in the context of the achieved computing power, memory flow rate, speed of matrix multiplication and the possibility of processing the image from the camera in real-time. The obta ined results were presented in graphical format, described and commented. We also provide an insight on applications that use this API for teaching deep neural networks, image processing, etc.

Autorzy

• dr inż. Przemysław Falkowski-Gilski link otwiera się w nowej karcie ,
• Maciej Plewka link otwiera się w nowej karcie