TITLE
LITERATURE REVIEW OF DEEP NEURAL NETWORKS FOR OBJECT TRACKING IN VIDEO FRAMES
AUTHOR(S)
Tsvetislava Lavchieva
ABSTRACT
This research explores the application of deep neural networks for object tracking in video frames. The main objective is to analyze and compare state-of-the-art multi-object tracking algorithms such as DeepSORT, ByteTrack, and FairMOT in terms of their accuracy, speed, and efficiency across various video scenarios. The study involves key stages including object detection, trajectory association, and evaluation of tracking stability under challenging conditions such as illumination changes, camera motion, and object occlusions. The study highlights the potential of deep learning-based tracking methods for practical applications in video surveillance, intelligent transportation systems, autonomous systems, and behavioral analysis. Keywords: deep neural networks, object tracking algorithms, artificial intelligence.
DOI
How to cite this article:
Tsvetislava Lavchieva, LITERATURE REVIEW OF DEEP NEURAL NETWORKS FOR OBJECT TRACKING IN VIDEO FRAMES, UNITECH – SELECTED PAPERS - 2025