Publicación:
Hand pose estimation from depth data with Convolutional Neural Networks

Vista sencilla de ítem
dc.contributor.authorVinagre Ruiz, Manuel
dc.date.accessioned2024-05-20T12:34:30Z
dc.date.available2024-05-20T12:34:30Z
dc.date.issued2017-09-02
dc.description.abstractThe estimation of hand position and orientation, pose, is of special interest in many applications related to Human Robot Interaction, such as human activity recognition, sign language interpretation, or as a human computer interface in virtual reality systems, advanced entertainment games, gesture-driven interfaces, and in teleoperated or in autonomous robotic systems. This project focusses on the problem of hand pose estimation using convolutional neural networks (CNN) from depth data. Recently, dierent CNN architectures have been proposed in order to nd an ecent and reliable methodolgy to resolve the complexity that involves the variablity in apperance of a hand, with its gestures, changes of orientation, occlusions and so. The use of CNN opens new opportunities for improvements in this research by providing the capability of learning from many samples. This work pretends to advance a step further on the hand pose estimation problem. With this aim, the hand pose estimation using CNN by modifying the output data with a new representation of the hand pose is proposed. An Euclidean Matrix Distance (EDM) is proposed as a hand pose representation. This representation encodes structural information of the hand pose and captures local correlations and dependecies between some hand keypoints. To evaluate the performance of the proposed method, dierent CNN architectures using EDM representation are explored and compared with the hand pose representation dened by the position of hand keypoints in the 3D Cartesian coordinate system. Experimental results show that using EDM representation as target layer in the convolutional network increases the performance of all the proposed architectures in terms of mean square error, both in training and testing sets. As a conclusion, this work shows that EDM representations help to reduce some ambiguities of current hand pose estimation methods using a CNN, by incorporating structural information of the hand and capturing keypoint/joint correlations. This research also gives some insights to investigate in future advances in hand pose estimation using CNN models and will help to explore new strategies for this problem.en
dc.description.versionversión final
dc.identifier.urihttps://hdl.handle.net/20.500.14468/14541
dc.language.isoen
dc.publisherUniversidad Nacional de Educación a Distancia (España). Escuela Técnica Superior de Ingeniería Informática. Departamento de Inteligencia Artificial
dc.relation.centerFacultades y escuelas::E.T.S. de Ingeniería Informática
dc.relation.degreeMáster Universitario en I.A. Avanzada: Fundamentos, Métodos y Aplicaciones
dc.relation.departmentInteligencia Artificial
dc.rightsinfo:eu-repo/semantics/openAccess
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/4.0/deed.es
dc.titleHand pose estimation from depth data with Convolutional Neural Networkses
dc.typetesis de maestríaes
dc.typemaster thesisen
dspace.entity.typePublication
Archivos
Bloque original
Mostrando 1 - 1 de 1
Miniatura
Nombre:
Vinagre_Ruiz_Manuel_TFM.pdf
Tamaño:
1.35 MB
Formato:
Adobe Portable Document Format
Descargar
Colecciones
Trabajos de fin de máster (TFM)