Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.11851/5500
Title: A fast deep learning based approach for basketball video analysis
Authors: Teket, O. M.
Yetik, İmam Şamil
Keywords: Deep learning
Re-identification
Video analysis
Issue Date: 2020
Publisher: Association for Computing Machinery
Source: 4th International Conference on Vision, Image and Signal Processing, ICVISP 2020, 9 December 2020 through 11 December 2020, , 167661
Abstract: The aim of this study is to obtain an algorithm to analyze a basketball training in real time. That is, the approach should be able to detect the correct players and scores for each player. For this purpose, we first propose a method to detect the shots and if they are makes or misses. Specifically, we use a deep learning model to detect the position of the hoop and background subtraction to detect the shot. Then, scoring is determined by another neural network using classification. After the determination of the shot, we go back in our buffered images to find the player who sent the shot. The player is detected by YOLOv3-tiny and identified by a deep unsupervised reidentification model. To keep the real-time aim, all models employ mobile networks as base with different parameters and training methods. The experiments were conducted on two training videos. The experiment results show the effectiveness of the current method with over 95% accuracy on scoring identification and up to 91.4% overall accuracy on reidentification of 4 players in real-time. © 2020 ACM.
URI: https://doi.org/10.1145/3448823.3448882
https://hdl.handle.net/20.500.11851/5500
ISBN: 9781450389532
Appears in Collections:Elektrik ve Elektronik Mühendisliği Bölümü / Department of Electrical & Electronics Engineering
Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collection

Show full item record

CORE Recommender

Page view(s)

12
checked on Dec 26, 2022

Google ScholarTM

Check

Altmetric


Items in GCRIS Repository are protected by copyright, with all rights reserved, unless otherwise indicated.