Paper Info
Title
Optimizing Through Learned Errors for Accurate Sports Field Registration
Abstract
We propose an optimization-based framework to register sports field templates onto broadcast videos. For accurate registration we go beyond the prevalent feed-forward paradigm. Instead, we propose to train a deep network that regresses the registration error, and then register images by finding the registration parameters that minimize the regressed error. We demonstrate the effectiveness of our method by applying it to real-world sports broadcast videos, outperforming the state of the art. We further apply our method on a synthetic toy example and demonstrate that our method brings significant gains even when the problem is simplified and unlimited training data is available <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> .
Year
DOI
Venue
2020
10.1109/WACV45572.2020.9093581
2020 IEEE Winter Conference on Applications of Computer Vision (WACV)
Keywords
DocType
ISSN
learned errors,accurate sports field registration,optimization-based framework,deep network,registration error,registration parameters,regressed error,real-world sports broadcast videos,feed-forward paradigm
Conference
2472-6737
ISBN
Citations 
PageRank 
978-1-7281-6554-7
1
0.35
References 
Authors
34
6
Name
Order
Citations
PageRank
Wei Jiang162.08
Higuera Juan Camilo Gamboa210.35
Baptiste Angles322.39
Weiwei Sun4132.31
Mehrsan Javan Roshtkhari525214.37
Kwang Moo Yi627116.65