Institutional Repository of Vilnius University
Dalinai prižiūrimas mokymas, skirtas veiksmų atpažinimui /
Abstract
dc:descriptionOne of the popular tasks of artificial neural networks is the recognition of actions in video. Despite the complexity of the task itself, the problem is also finding, preparing and processing data used for learning procedures. Because this way of using neural networks is quite new, there is not much prepared data sets for network training, which means that all data sets have to be gathered using human aid, which takes a lot of time. This work introduces the recognition of actions in video clips using neural networks. Further, a problem of data acquisition is being analyzed, by creating a software which reduces human contribution in finding and processing data for network training. And finally, using created program there are experiments, which tests automatic data acquisition in practice.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2019
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Valiauga, Arnas,
- Contributors dc:contributor
-
- Grevys, Andrius
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD81918382&prefLang=en_US