Classification of occluded 2d objects using deep learning of 3d shape surfaces

Vasileios Tzitzilonis, George Apostolopoulos, Vassilios Kappatos, Evangelos Dermatas

Publikation: Kapitel i bog/rapport/konference-proceedingKonferencebidrag i proceedingsForskningpeer review

Abstrakt

This paper presents a novel deep learning method for partially occluded 2D object classification. A 2D Convolutional Neural Network (CNN) was trained with partial and whole images of the 3D models obtained from different camera views. The efficiency of the proposed method in classifying partial objects in 40 categories is more than 80% in most objects and exceeds 95% in some of them.

OriginalsprogEngelsk
TitelProceedings - 10th Hellenic Conference on Artificial Intelligence, SETN 2018
ForlagAssociation for Computing Machinery / Special Interest Group on Programming Languages
Publikationsdato9. jul. 2018
ISBN (Elektronisk)978-1-4503-6433-1
DOI
StatusUdgivet - 9. jul. 2018
Begivenhed10th Hellenic Conference on Artificial Intelligence, SETN 2018 - Patras, Grækenland
Varighed: 9. jul. 201812. jul. 2018

Konference

Konference10th Hellenic Conference on Artificial Intelligence, SETN 2018
LandGrækenland
ByPatras
Periode09/07/201812/07/2018
SponsorHellenic Artificial Intelligence Society (EETN), University of Patras, University of Thessaly

Fingeraftryk Dyk ned i forskningsemnerne om 'Classification of occluded 2d objects using deep learning of 3d shape surfaces'. Sammen danner de et unikt fingeraftryk.

Citationsformater