Semantic Scholar Open Access 2022 46 sitasi

Tuning of data augmentation hyperparameters in deep learning to building construction image classification with small datasets

A. L. Ottoni Raphael M. de Amorim M. Novo D. Costa

Abstrak

Deep Learning methods have important applications in the building construction image classification field. One challenge of this application is Convolutional Neural Networks adoption in a small datasets. This paper proposes a rigorous methodology for tuning of Data Augmentation hyperparameters in Deep Learning to building construction image classification, especially to vegetation recognition in facades and roofs structure analysis. In order to do that, Logistic Regression models were used to analyze the performance of Convolutional Neural Networks trained from 128 combinations of transformations in the images. Experiments were carried out with three architectures of Deep Learning from the literature using the Keras library. The results show that the recommended configuration (Height Shift Range = 0.2; Width Shift Range = 0.2; Zoom Range =0.2) reached an accuracy of $$95.6\%$$ 95.6 % in the test step of first case study. In addition, the hyperparameters recommended by proposed method also achieved the best test results for second case study: $$93.3\%$$ 93.3 % .

Penulis (4)

A

A. L. Ottoni

R

Raphael M. de Amorim

M

M. Novo

D

D. Costa

Format Sitasi

Ottoni, A.L., Amorim, R.M.d., Novo, M., Costa, D. (2022). Tuning of data augmentation hyperparameters in deep learning to building construction image classification with small datasets. https://doi.org/10.1007/s13042-022-01555-1

Akses Cepat

Lihat di Sumber doi.org/10.1007/s13042-022-01555-1
Informasi Jurnal
Tahun Terbit
2022
Bahasa
en
Total Sitasi
46×
Sumber Database
Semantic Scholar
DOI
10.1007/s13042-022-01555-1
Akses
Open Access ✓