MOBILE OBJECT DETECTION ALGHORITHMS USING TENSAFLOW

ЧАСТЬ 1 ТОМ 2 ИЮЛЬ 2023 год

Abstract

In recent years, deep learning has advanced, enabling us to
build sophisticated machine learning models for object detection in photos
regardless of the properties of the objects to be recognized. This advancement has
made it possible for engineers to replace current heuristics -based systems with
machine learning models that perform better. In this study, we assess the
effectiveness of utilizing deep learning models as feature extractors for existing
algorithms or end-to-end systems for object detection in real-time video feeds on
mobile devices in terms of object detection performance and inference delay. In
compared to existing techniques, our results demonstrate a considerable
improvement in object detection performance when transfer learning is applied to
neural networks that have been optimized for mobile application