A Genetic Programming Approach to Integrate Multilayer CNN Features for Image Classification

Wei Ta Chu, Hao An Chu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Fusing information extracted from multiple layers of a convolutional neural network has been proven effective in several domains. Common fusion techniques include feature concatenation and Fisher embedding. In this work, we propose to fuse multilayer information by genetic programming (GP). With the evolutionary strategy, we iteratively fuse multilayer information in a systematic manner. In the evaluation, we verify the effectiveness of discovered GP-based representations on three image classification datasets, and discuss characteristics of the GP process. This study is one of the few works to fuse multilayer information based on an evolutionary strategy. The reported preliminary results not only demonstrate the potential of the GP fusion scheme, but also inspire future study in several aspects.

Original languageEnglish
Title of host publicationMultiMedia Modeling - 25th International Conference, MMM 2019, Proceedings
EditorsIoannis Kompatsiaris, Stefanos Vrochidis, Vasileios Mezaris, Wen-Huang Cheng, Benoit Huet, Cathal Gurrin
PublisherSpringer Verlag
Pages640-651
Number of pages12
ISBN (Print)9783030057091
DOIs
Publication statusPublished - 2019 Jan 1
Event25th International Conference on MultiMedia Modeling, MMM 2019 - Thessaloniki, Greece
Duration: 2019 Jan 82019 Jan 11

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11295 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference25th International Conference on MultiMedia Modeling, MMM 2019
CountryGreece
CityThessaloniki
Period19-01-0819-01-11

Fingerprint

Genetic programming
Image classification
Image Classification
Genetic Programming
Multilayer
Multilayers
Electric fuses
Integrate
Evolutionary Strategy
Fusion
Fusion reactions
Concatenation
Neural Networks
Verify
Neural networks
Evaluation
Demonstrate

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Chu, W. T., & Chu, H. A. (2019). A Genetic Programming Approach to Integrate Multilayer CNN Features for Image Classification. In I. Kompatsiaris, S. Vrochidis, V. Mezaris, W-H. Cheng, B. Huet, & C. Gurrin (Eds.), MultiMedia Modeling - 25th International Conference, MMM 2019, Proceedings (pp. 640-651). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11295 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-05710-7_53
Chu, Wei Ta ; Chu, Hao An. / A Genetic Programming Approach to Integrate Multilayer CNN Features for Image Classification. MultiMedia Modeling - 25th International Conference, MMM 2019, Proceedings. editor / Ioannis Kompatsiaris ; Stefanos Vrochidis ; Vasileios Mezaris ; Wen-Huang Cheng ; Benoit Huet ; Cathal Gurrin. Springer Verlag, 2019. pp. 640-651 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{f088f7b4dff1457a90c5b6820c3d6641,
title = "A Genetic Programming Approach to Integrate Multilayer CNN Features for Image Classification",
abstract = "Fusing information extracted from multiple layers of a convolutional neural network has been proven effective in several domains. Common fusion techniques include feature concatenation and Fisher embedding. In this work, we propose to fuse multilayer information by genetic programming (GP). With the evolutionary strategy, we iteratively fuse multilayer information in a systematic manner. In the evaluation, we verify the effectiveness of discovered GP-based representations on three image classification datasets, and discuss characteristics of the GP process. This study is one of the few works to fuse multilayer information based on an evolutionary strategy. The reported preliminary results not only demonstrate the potential of the GP fusion scheme, but also inspire future study in several aspects.",
author = "Chu, {Wei Ta} and Chu, {Hao An}",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/978-3-030-05710-7_53",
language = "English",
isbn = "9783030057091",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "640--651",
editor = "Ioannis Kompatsiaris and Stefanos Vrochidis and Vasileios Mezaris and Wen-Huang Cheng and Benoit Huet and Cathal Gurrin",
booktitle = "MultiMedia Modeling - 25th International Conference, MMM 2019, Proceedings",
address = "Germany",

}

Chu, WT & Chu, HA 2019, A Genetic Programming Approach to Integrate Multilayer CNN Features for Image Classification. in I Kompatsiaris, S Vrochidis, V Mezaris, W-H Cheng, B Huet & C Gurrin (eds), MultiMedia Modeling - 25th International Conference, MMM 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11295 LNCS, Springer Verlag, pp. 640-651, 25th International Conference on MultiMedia Modeling, MMM 2019, Thessaloniki, Greece, 19-01-08. https://doi.org/10.1007/978-3-030-05710-7_53

A Genetic Programming Approach to Integrate Multilayer CNN Features for Image Classification. / Chu, Wei Ta; Chu, Hao An.

MultiMedia Modeling - 25th International Conference, MMM 2019, Proceedings. ed. / Ioannis Kompatsiaris; Stefanos Vrochidis; Vasileios Mezaris; Wen-Huang Cheng; Benoit Huet; Cathal Gurrin. Springer Verlag, 2019. p. 640-651 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11295 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - A Genetic Programming Approach to Integrate Multilayer CNN Features for Image Classification

AU - Chu, Wei Ta

AU - Chu, Hao An

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Fusing information extracted from multiple layers of a convolutional neural network has been proven effective in several domains. Common fusion techniques include feature concatenation and Fisher embedding. In this work, we propose to fuse multilayer information by genetic programming (GP). With the evolutionary strategy, we iteratively fuse multilayer information in a systematic manner. In the evaluation, we verify the effectiveness of discovered GP-based representations on three image classification datasets, and discuss characteristics of the GP process. This study is one of the few works to fuse multilayer information based on an evolutionary strategy. The reported preliminary results not only demonstrate the potential of the GP fusion scheme, but also inspire future study in several aspects.

AB - Fusing information extracted from multiple layers of a convolutional neural network has been proven effective in several domains. Common fusion techniques include feature concatenation and Fisher embedding. In this work, we propose to fuse multilayer information by genetic programming (GP). With the evolutionary strategy, we iteratively fuse multilayer information in a systematic manner. In the evaluation, we verify the effectiveness of discovered GP-based representations on three image classification datasets, and discuss characteristics of the GP process. This study is one of the few works to fuse multilayer information based on an evolutionary strategy. The reported preliminary results not only demonstrate the potential of the GP fusion scheme, but also inspire future study in several aspects.

UR - http://www.scopus.com/inward/record.url?scp=85059856369&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85059856369&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-05710-7_53

DO - 10.1007/978-3-030-05710-7_53

M3 - Conference contribution

AN - SCOPUS:85059856369

SN - 9783030057091

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 640

EP - 651

BT - MultiMedia Modeling - 25th International Conference, MMM 2019, Proceedings

A2 - Kompatsiaris, Ioannis

A2 - Vrochidis, Stefanos

A2 - Mezaris, Vasileios

A2 - Cheng, Wen-Huang

A2 - Huet, Benoit

A2 - Gurrin, Cathal

PB - Springer Verlag

ER -

Chu WT, Chu HA. A Genetic Programming Approach to Integrate Multilayer CNN Features for Image Classification. In Kompatsiaris I, Vrochidis S, Mezaris V, Cheng W-H, Huet B, Gurrin C, editors, MultiMedia Modeling - 25th International Conference, MMM 2019, Proceedings. Springer Verlag. 2019. p. 640-651. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-05710-7_53