Data Augmentation and Convolutional Network Architecture Influence on Distributed Learning

Victor Forattini Jansen; Emanuel Teixeira Martins; Yasmin Souza Lima; Flavio de Oliveira Silva; Rodrigo Moreira; Larissa Ferreira Rodrigues Moreira

doi:10.22456/2175-2745.143508

Authors

Victor Forattini Jansen Universidade Federal de Viçosa (UFV) https://orcid.org/0009-0005-8179-1717
Emanuel Teixeira Martins Universidade Federal de Viçosa (UFV) https://orcid.org/0009-0005-5528-5036
Yasmin Souza Lima Universidade Federal de Viçosa (UFV) https://orcid.org/0009-0000-4963-4440
Flavio de Oliveira Silva Department of Informatics University of Minho (UMINHO) & Federal University of Uberlândia (UFU) https://orcid.org/0000-0001-7051-7396
Rodrigo Moreira Universidade Federal de Viçosa (UFV) https://orcid.org/0000-0002-9328-8618
Larissa Ferreira Rodrigues Moreira Universidade Federal de Viçosa (UFV) https://orcid.org/0000-0001-8947-9182

DOI:

https://doi.org/10.22456/2175-2745.143508

Keywords:

distributed learning, rice classification, data augumentation, CNN, factorial design

Abstract

Convolutional Neural Networks (CNNs) have proven to be highly effective in solving a broad spectrum of computer vision tasks, such as classification, identification, and segmentation. These methods can be deployed in both centralized and distributed environments, depending on the computational demands of the task. While much of the literature has focused on the explainability of CNNs, which is essential for building trust and confidence in their predictions, there remains a gap in understanding their impact on computational resources, particularly in distributed training contexts. In this study, we analyze how CNN architectures primarily influence model accuracy and investigate additional factors that affect computational efficiency in distributed systems. Our findings contribute valuable insights for optimizing the deployment of CNNs in resource-intensive scenarios, paving the way for further exploration of variables critical to distributed learning.

Downloads

Download data is not yet available.

References

SONG, J. et al. Dimo-cnn: Deep learning toolkit-accelerated analytical modeling and optimization of cnn hardware and dataflow. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, p. 1–1, 2024.

Rodrigues Moreira, L. F. et al. An Artificial Intelligence-as-a-Service Architecture for deep learning model embodiment on low-cost devices: A case study of COVID-19 diagnosis. Applied Soft Computing, v. 134, p. 110014, 2023. ISSN 1568-4946.

WIEDEMANN, S.; MÜLLER, K.-R.; SAMEK, W. Compact and computationally efficient representation of deep neural networks. IEEE Transactions on Neural Networks and Learning Systems, v. 31, n. 3, p. 772–785, 2020.

PETCHIAMMAL et al. Paddy doctor: A visual image dataset for automated paddy disease classification and benchmarking. In: Proceedings of the 6th Joint International Conference on Data Science & Management of Data (10th ACM IKDD CODS and 28th COMAD). New York, NY, USA: Association for Computing Machinery, 2023. (CODS-COMAD ’23), p. 203–207. ISBN 9781450397971. Disponível em: https://doi.org/10.1145/3570991.3570994.

COSTA, F. et al. Self-portrait to ID document face matching: CNN-based face verification in cross-domain scenario. In: Anais do XVII Workshop de Visão Computacional. Porto Alegre, RS, Brasil: SBC, 2021. p. 31–36. ISSN 0000-0000. Disponível em: https://sol.sbc.org.br/index.php/wvc/article/view/18885.

ROCHA, E. L. da; RODRIGUES, L.; MARI, J. F. Maize leaf disease classification using convolutional neural networks and hyperparameter optimization. In: SBC. Anais do XVI Workshop de Visão Computacional. [S.l.], 2020. p. 104–110.

BOUZON, M. et al. Automatic kidney stone detection using low-cost CNN with coronal CT images. In: Anais do XVIII Workshop de Visão Computacional. Porto Alegre, RS, Brasil: SBC, 2023. p. 24–29. ISSN 0000-0000. Disponível em: https://sol.sbc.org.br/index.php/wvc/article/view/27527.

MOREIRA, R. et al. Agrolens: A low-cost and green-friendly smart farm architecture to support real-time leaf disease diagnostics. Internet of Things, v. 19, p. 100570, 2022. ISSN 2542-6605.

PARK, S.; LEE, J.; KIM, H. Hardware resource analysis in distributed training with edge devices. Electronics, v. 9, n. 1, 2020. ISSN 2079-9292. Disponível em: https://www.mdpi.com/2079-9292/9/1/28.

MOHAIDAT, T.; KHALIL, K. A survey on neural network hardware accelerators. IEEE Transactions on Artificial Intelligence, v. 5, n. 8, p. 3801–3822, 2024.

AGGARWAL, M. et al. Federated transfer learning for rice-leaf disease classification across multiclient cross-silo datasets. Agronomy, v. 13, n. 10, 2023. ISSN 2073-4395. Disponível em: https://www.mdpi.com/2073-4395/13/10/2483.

NI, H. et al. Classification of typical pests and diseases of rice based on the ECA attention mechanism. Agriculture, v. 13, n. 5, 2023. ISSN 2077-0472. Disponível em: https://www.mdpi.com/2077-0472/13/5/1066.

YANG, Y. et al. A lightweight rice disease identification network based on attention mechanism and dynamic convolution. Ecological Informatics, v. 78, p. 102320, 2023. ISSN 1574-9541. Disponível em: https://www.sciencedirect.com/science/article/pii/S1574954123003497.

SETHY, P. K. et al. Deep feature based rice leaf disease identification using support vector machine. Computers and Electronics in Agriculture, v. 175, p. 105527, 2020. ISSN 0168-1699. Disponível em: https://www.sciencedirect.com/science/article/pii/S0168169919326997.