Wyniki wyszukiwania - BazTech

1

Single target tracking algorithm for lightweight Siamese networks based on global attention

Wang Zhentao, He Xiaowei, Cheng Rao

Bulletin of the Polish Academy of Sciences. Technical Sciences

|

2022

|

Vol. 70, nr 3

art. no. e139961

EN

Object tracking based on Siamese networks has achieved great success in recent years, but increasingly advanced trackers are also becoming cumbersome, which will severely limit deployment on resource-constrained devices. To solve the above problems, we designed a network with the same or higher tracking performance as other lightweight models based on the SiamFC lightweight tracking model. At the same time, for the problems that the SiamFC tracking network is poor in processing similar semantic information, deformation, illumination change, and scale change, we propose a global attention module and different scale training and testing strategies to solve them. To verify the effectiveness of the proposed algorithm, this paper has done comparative experiments on the ILSVRC, OTB100, VOT2018 datasets. The experimental results show that the method proposed in this paper can significantly improve the performance of the benchmark algorithm.

2

Domains as Models for Semantic Information

Muravitsky A. Y.

Fundamenta Informaticae

|

2015

|

Vol. 138, nr 1/2

207--225

EN

We propose a framework in terms of domain theory for semantic informationmodels. We show how an artificial agent (the computer) can operate within such a model in a multiple attitude environment (fuzziness) where information is conveyed. We illustrate our approach by two examples — taking as the set of the degrees of reliability Kleene’s 3-valued strong logic and Belnap-Dunn’s 4-valued logic.

3

System nawigacyjny wykorzystujący informację semantyczną

Siemiątkowska B., Szklarski J., Gnatowski M., Borkowski A.

Prace Naukowe Politechniki Warszawskiej. Elektronika

|

2010

|

z. 175, t. 2

645-654

PL

W poniższej pracy przedstawiono system nawigacyjny robota mobilnego. W procesie planowania trasy wykorzystuje się semantyczną wiedzę o otoczeniu. Robot wyposażony w skaner laserowy 3D analizuje otoczenie i przypisuje obserwowanym obiektom etykiety. Cel do którego robot ma dotrzeć jest wskazywany poprzez podanie nazwy obiektu. Możemy więc wydać polecenie typu jedź do ściany, do drzwi, czy też umywalki. Zastosowano hybrydową rastrowo-obiektową reprezentację otoczenia. W procesie planowania trasy zastosowano sieci komórkowe.

EN

In this article we present a system which allows a mobile robot to navigate in an outdoor or indoor environment. Data obtained from a 3D laser range finder is analyzed and semantic labels are attached to the detected objects. The dual grid based and semantic map is built. The obstacle-free path is generated using a Cellular Neural Network. The goal for the robot is given using semantic labels. When the same label is attached to many objects the cheapest path is found. The path planning method is fast and allows taking into account various features of the environment and types of robots. In comparison to the potential field method the algorithm proposed in this paper does not suffer from local minima problem. The experiments were performed in real indoor and outdoor environments.

4

Semantic Sparse Representation of Disease Patterns

Przelaskowski A.

International Journal of Electronics and Telecommunications

|

2010

|

Vol. 56, No. 3

273-280

EN

Sparse data representation is discussed in a context of useful fundamentals led to semantic content description and extraction of information. Disease patterns as semantic information extracted from medical images were underlined because of discussed application of computer-aided diagnosis. Compressive sensing rules were adjusted to the requirements of diagnostic pattern recognition. Proposed methodology of sparse disease patterns considers accuracy of sparse representation to estimate target content for detailed analysis. Semantics of sparse representation were modeled by morphological content analysis. Subtle or hidden components were extracted and displayed to increase information completeness. Usefulness of sparsity was verified for computer-aided diagnosis of stroke based on brain CT scans. Implemented method was based on selective and sparse representation of subtle hypodensity to improve diagnosis. Visual expression of disease signatures was fixed to radiologist requirements, domain knowledge and experimental analysis issues. Diagnosis assistance suitability was proven by experimental subjective rating and automatic recognition.

5

Semantic information within the BEATCA framework

Ciesielski K., Czerski D., Dramiński M., Kłopotek M. A., Wierzchoń B.

Control and Cybernetics

|

2010

|

Vol. 39, no 2

377-399

EN

In this paper we investigate the impact of semantic information on the quality of hierarchical, fuzzy-based clustering of a collection of textual documents. We show that via a relevant tagging of a part of the documents one can improve the quality of overall clustering, both of tagged and un-tagged documents.