Semi-supervised feature sharing for efficient video segmentation

Ponomarchuk, Anton

Home
→
Students Research & Project Works | Роботи студентів
→
Факультет прикладних наук
→
Освітня програма наук про дані
→
2019
→
View Item

dc.contributor.author	Ponomarchuk, Anton
dc.date.accessioned	2019-02-19T15:35:52Z
dc.date.available	2019-02-19T15:35:52Z
dc.date.issued	2019
dc.identifier.citation	Ponomarchuk, Anton. Semi-supervised feature sharing for efficient video segmentation : Master Thesis : manuscript / Anton Ponomarchuk ; Supervisor Andrey Luzan ; Ukrainian Catholic University, Department of Computer Sciences. – Lviv : [s.n.], 2019. – 28 p. : ill.	uk
dc.identifier.uri	http://er.ucu.edu.ua/handle/1/1335
dc.language.iso	en	uk
dc.subject	Semi-supervised feature sharing	uk
dc.subject	Video semantic segmentation	uk
dc.subject	Loss function and accuracy	uk
dc.title	Semi-supervised feature sharing for efficient video segmentation	uk
dc.type	Preprint	uk
dc.status	Публікується вперше	uk
dc.description.abstracten	In robot sensing and automotive driving domains, producing precise semantic segmentation masks for images can help greatly with environment understanding and, as a result, better interaction with it. These tasks usually need to be processed for images with more the 2 object’s classes. Moreover, semantic segmentation should be done for a short period. Almost all approaches that try to solve this task used heavyweight end-to-end deep neural network or external blocks like GRU [14], LSTM[25] or optical flow [1]. In this work, we provide a deep neural network architecture for learning to extract global high-level features and propagate them among the images that describe the same video’s scene, for speeding up image processing. We provide a propagation strategy without any external blocks. We also provide loss function for training such network with the dataset, where the vast number of images don’t have a segmentation mask.	uk