We demonstrate in our paper, an implementation on Microsoft HoloLens, deep learning supported in the context of object detection. The main aim of this system is to create the more accurate object detection model for Augmented Reality using communication between the deep learning processing and the Microsoft HoloLens as Input/Output device. This system aims to help the wearable device user to detect and to recognize between objects in real world. For the object detection approach, a deep learning model has been used for the implementation of this system called YOLO. This model is near to real-time and it supports to detect more than 9000 objects. Our system provides the annotation of augmented object detected and its limitation area or bounding box via HoloLens. It allows to detect the new position of moving object in a few milliseconds. Preliminary results show a great rate of object detection with a detection time comparable.
Paper
Full text
Accurate Object Detection System on HoloLens Using YOLO Algorithm
Semantic Scholar · Computer Science · 2019
Abstract
We demonstrate in our paper, an implementation on Microsoft HoloLens, deep learning supported in the context of object detection. The main aim of this system is to create the more accurate object detection model for Augmented Reality using communication between the deep learning processing and the Microsoft HoloLens as Input/Output device. This system aims to help the wearable device user to detect and to recognize between objects in real world. For the object detection approach, a deep learning model has been used for the implementation of this system called YOLO. This model is near to real-time and it supports to detect more than 9000 objects. Our system provides the annotation of augmented object detected and its limitation area or bounding box via HoloLens. It allows to detect the new position of moving object in a few milliseconds. Preliminary results show a great rate of object detection with a detection time comparable.