A Classification of Golek Menak Dancer Poses Based on Learning Vector Quantization (LVQ) and Genetic Algorith

Joko, Sutopo (2018) A Classification of Golek Menak Dancer Poses Based on Learning Vector Quantization (LVQ) and Genetic Algorith. A Classification of Golek Menak Dancer Poses Based on Learning Vector Quantization (LVQ) and Genetic Algorith, 13 (6). pp. 1-7. ISSN 1816-949X

Text 1. A classification of Golek Menak Dancer poses based on Learning Vector Quantization (LVQ) and genetic algorithm.pdf Download (203kB)

Abstract

Abstract : There are still rarely discussed the Golek Menak dance from technology perspective, especially in motion capture detection. Our study proposed a classification model using Learning Vector Quantization ( LVQ ) which combined with Genetic Algorithm (GA). This is a novelty that the author considered important to improve the accuracy in detecting Golek Menak dancer and resolve their complexity through tensor rule of Canonical Parafac - Alternating Least Square (CP-ALS) method. We also have taken eight poses representing Golek Menak dancer poses and estimating their moves in geometric shapes that cause translational, rotational, dilatational, reflection and geometric slope (shear) translations. The tensor rule is important in our study to estimate the geometrical transformation model of the dancer (e.g.,body,hand,head,leg and time duration). After LVQ is implemented, we can finally, deleting repeated poses into single pose as standardized poses. The tensor rule also can reduce the impact on kinematic transformation. Whereas genetic algorithm will find the value of fitness, the higher value of the joints, the more likely the joints to represent the dancer poses. Finally, we presented the result of the body transformation of the dancer motion with complete combination of CP, LVQ and GA to provide a model with higher accuracy . Our study brings contribution to expand the theory of CP, LVQ and GA in the dance motion recognition.

Item Type:	Article
Subjects:	N Fine Arts > NX Arts in general
Divisions:	Fakultas Sains Dan Teknologi > Data Science
Depositing User:	ST., MT. Joko Sutopo
Date Deposited:	04 Apr 2023 09:48
Last Modified:	04 Apr 2023 09:48
URI:	http://eprints.uty.ac.id/id/eprint/12560

Actions (login required)

View Item