Application of machine learning models for 3D object recognition for iOS
Students Name: Kolisnyk Andrii Ihorovych
Qualification Level: magister
Speciality: Information Technology Design
Institute: Institute of Computer Science and Information Technologies
Mode of Study: full
Academic Year: 2022-2023 н.р.
Language of Defence: ukrainian
Abstract: Kolisnyk A.I., Beley O.I. Research of machine learning models for 3D object recognition for IOS. Master’s qualification work. - Lviv Polytechnic National University, Lviv, 2022. Extended abstract. Master’s qualification work is devoted to the development of a mobile application for recognizing and generating 3D objects for iOS. In the first section, the features of iOS application development were presented. General characteristics of object recognition and principles of computer vision. The second part of the section presents the methodology of object recognition and the architecture of the application. The second section describes two main families for object detection. A detailed analysis of each family is carried out and the YOLO algorithm is selected for object detection. The second part of the section describes the algorithm for object scanning. It highlighted how the object can be scanned. The third section presents the architecture of the application and the coordinator for navigation. The creation of machine learning technologies for 3D object recognition using the Xcode development environment is carried out. The second part of the section is responsible for the implementation of technologies for detecting and scanning 3D objects [2] and demonstrates the testing of the algorithm and the result of the program on a real object. The object of research is the processes of object recognition in iOS. The subject of research - machine modelling of processes of recognition and scanning of 3D objects for iOS. Purpose - to study the models of detection and scanning of 3D objects for iOS. The result of the work is a developed mobile application with which the user can easily scan any real object and export it for further use. The total volume of work is 94 pages, with 43 images, and 14 references. Keywords - machine learning, recognition and scanning of 3D objects, modelling, machine vision. List of used literature sources: 1. Machine learning on mobile: What can you actually do with it? [Electronic resource]. – Access mode: https://heartbeat.comet.ml/machine-learning-on-mobile-what-can-you-actually-do-with-it-8437fa782165 2. How to make 3D models from a single image [Electronic resource]. – Access mode: https://medium.com/mlearning-ai/how-to-make-3d-models-from-a-single-image-d0eccc9209ba