On Salient Object Detection

dc.contributor.author	Prots, Olha
dc.date.accessioned	2024-02-14T09:27:10Z
dc.date.available	2024-02-14T09:27:10Z
dc.date.issued	2022
dc.identifier.citation	Prots, Olha. On Salient Object Detection / Prots, Olha; Supervisor: Mr. Orest Kupyn; Ukrainian Catholic University, Department of Computer Sciences. – Lviv: 2022. – 37 p.	uk
dc.identifier.uri	https://er.ucu.edu.ua/handle/1/4412
dc.language.iso	en	uk
dc.title	On Salient Object Detection	uk
dc.type	Preprint	uk
dc.status	Публікується вперше	uk
dc.description.abstracten	Salient object detection is the process of finding the most visually catching objects in the image. This can be very beneficial to outline the region of interest and use that information for further image processing. In this paper, we review the most common approaches to this problem and propose a simple approach that strives to be compact and efficient. Since most SOTA solutions achieve their accuracy by sacrificing computational efficiency, they are not suitable for limited resources. We test an approach that achieves comparative results on much smaller UNet and Unet++ models.	uk