Non-rigid 3D shape classification based on convolutional neural networks

Jan Franco Llerena Quenaya; Cristian José Lopez Del Alamo

doi:10.1109/LA-CCI.2017.8285693

Non-rigid 3D shape classification based on convolutional neural networks

Jan Franco Llerena Quenaya, Cristian José Lopez Del Alamo

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Over the years, the scientific interest towards 3D models analysis has become more popular. Problems such as classification, retrieval and matching are studied with the idea to offer robust solutions. This paper introduces a 3D object classification method for non-rigid shapes, based on the detection of key points, the use of spectral descriptors and deep learning techniques. We adopt an approach of converting the models into a "spectral image". By extracting interest points and calculating three types of spectral descriptors (HKS, WKS and GISIF), we generate a three-channel input to a convolutional neural network. This CNN is trained to automatically learn features such as topology of 3D models. The results are evaluated and analyzed using the Non-Rigid Classification Benchmark SHREC 2011. Our proposal shows promising results in classification tasks compared to other methods, and also it is robust under several types of transformations.

Original language	English
Title of host publication	2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1-6
Number of pages	6
ISBN (Electronic)	9781538637340
DOIs	https://doi.org/10.1109/LA-CCI.2017.8285693
State	Published - 7 Feb 2018
Externally published	Yes
Event	2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Arequipa, Peru Duration: 8 Nov 2017 → 10 Nov 2017

Publication series

Name	2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings
Volume	2017-November

Conference

Conference	2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017
Country/Territory	Peru
City	Arequipa
Period	8/11/17 → 10/11/17

Bibliographical note

Publisher Copyright:
© 2017 IEEE.

Access to Document

10.1109/LA-CCI.2017.8285693

Cite this

Quenaya, J. F. L., & Lopez Del Alamo, C. J. (2018). Non-rigid 3D shape classification based on convolutional neural networks. In 2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings (pp. 1-6). (2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings; Vol. 2017-November). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/LA-CCI.2017.8285693

Quenaya, Jan Franco Llerena ; Lopez Del Alamo, Cristian José. / Non-rigid 3D shape classification based on convolutional neural networks. 2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 1-6 (2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings).

@inproceedings{e73748819aea4759b42efb45e2aebc7e,

title = "Non-rigid 3D shape classification based on convolutional neural networks",

abstract = "Over the years, the scientific interest towards 3D models analysis has become more popular. Problems such as classification, retrieval and matching are studied with the idea to offer robust solutions. This paper introduces a 3D object classification method for non-rigid shapes, based on the detection of key points, the use of spectral descriptors and deep learning techniques. We adopt an approach of converting the models into a {"}spectral image{"}. By extracting interest points and calculating three types of spectral descriptors (HKS, WKS and GISIF), we generate a three-channel input to a convolutional neural network. This CNN is trained to automatically learn features such as topology of 3D models. The results are evaluated and analyzed using the Non-Rigid Classification Benchmark SHREC 2011. Our proposal shows promising results in classification tasks compared to other methods, and also it is robust under several types of transformations.",

author = "Quenaya, {Jan Franco Llerena} and {Lopez Del Alamo}, {Cristian Jos{\'e}}",

note = "Publisher Copyright: {\textcopyright} 2017 IEEE.; 2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 ; Conference date: 08-11-2017 Through 10-11-2017",

year = "2018",

month = feb,

day = "7",

doi = "10.1109/LA-CCI.2017.8285693",

language = "Ingl{\'e}s",

series = "2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "1--6",

booktitle = "2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings",

address = "Estados Unidos",

}

Quenaya, JFL & Lopez Del Alamo, CJ 2018, Non-rigid 3D shape classification based on convolutional neural networks. in 2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings. 2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings, vol. 2017-November, Institute of Electrical and Electronics Engineers Inc., pp. 1-6, 2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017, Arequipa, Peru, 8/11/17. https://doi.org/10.1109/LA-CCI.2017.8285693

Non-rigid 3D shape classification based on convolutional neural networks. / Quenaya, Jan Franco Llerena; Lopez Del Alamo, Cristian José.
2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2018. p. 1-6 (2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings; Vol. 2017-November).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Non-rigid 3D shape classification based on convolutional neural networks

AU - Quenaya, Jan Franco Llerena

AU - Lopez Del Alamo, Cristian José

PY - 2018/2/7

Y1 - 2018/2/7

N2 - Over the years, the scientific interest towards 3D models analysis has become more popular. Problems such as classification, retrieval and matching are studied with the idea to offer robust solutions. This paper introduces a 3D object classification method for non-rigid shapes, based on the detection of key points, the use of spectral descriptors and deep learning techniques. We adopt an approach of converting the models into a "spectral image". By extracting interest points and calculating three types of spectral descriptors (HKS, WKS and GISIF), we generate a three-channel input to a convolutional neural network. This CNN is trained to automatically learn features such as topology of 3D models. The results are evaluated and analyzed using the Non-Rigid Classification Benchmark SHREC 2011. Our proposal shows promising results in classification tasks compared to other methods, and also it is robust under several types of transformations.

AB - Over the years, the scientific interest towards 3D models analysis has become more popular. Problems such as classification, retrieval and matching are studied with the idea to offer robust solutions. This paper introduces a 3D object classification method for non-rigid shapes, based on the detection of key points, the use of spectral descriptors and deep learning techniques. We adopt an approach of converting the models into a "spectral image". By extracting interest points and calculating three types of spectral descriptors (HKS, WKS and GISIF), we generate a three-channel input to a convolutional neural network. This CNN is trained to automatically learn features such as topology of 3D models. The results are evaluated and analyzed using the Non-Rigid Classification Benchmark SHREC 2011. Our proposal shows promising results in classification tasks compared to other methods, and also it is robust under several types of transformations.

UR - http://www.scopus.com/inward/record.url?scp=85050367614&partnerID=8YFLogxK

U2 - 10.1109/LA-CCI.2017.8285693

DO - 10.1109/LA-CCI.2017.8285693

M3 - Contribución a la conferencia

AN - SCOPUS:85050367614

T3 - 2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings

SP - 1

EP - 6

BT - 2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017

Y2 - 8 November 2017 through 10 November 2017

ER -

Quenaya JFL, Lopez Del Alamo CJ. Non-rigid 3D shape classification based on convolutional neural networks. In 2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2018. p. 1-6. (2017 IEEE Latin American Conference on Computational Intelligence, LA-CCI 2017 - Proceedings). doi: 10.1109/LA-CCI.2017.8285693

Non-rigid 3D shape classification based on convolutional neural networks

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this