Predicting Protein Content in Grain Using Hyperspectral Deep Learning
Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt
Standard
Predicting Protein Content in Grain Using Hyperspectral Deep Learning. / Galbo Engstrøm1, Ole-Christian; Dreier, Erik Schou; Steenstrup Pedersen, Kim.
Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) . IEEE, 2021. s. 1372-1380.Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Predicting Protein Content in Grain Using Hyperspectral Deep Learning
AU - Galbo Engstrøm1,, Ole-Christian
AU - Dreier, Erik Schou
AU - Steenstrup Pedersen, Kim
PY - 2021
Y1 - 2021
N2 - We assess the possibility of performing regression analysis on hyperspectral images utilizing the entire spatio-spectral data cube in convolutional neural networks using protein regression analysis of bulk wheat grain kernels as a test case. By introducing novel modifications of the well-known convolutional neural network, ResNet-18, we are able to significantly increase its performance on hyperspectral images. Our modifications consist of firstly applying a 3D convolution layer enabling learning of spectral derivatives that 2D spatial convolution is unable to learn, and secondly, the application of a (1 x 1) 2D convolution layer that downsamples the spectral dimension. Analysis of the responses learned by the convolution kernels in our modifications reveals meaningful representations of the input data cube that reduce noise and enable the subsequent ResNet-18 to perform more accurate regression analysis.
AB - We assess the possibility of performing regression analysis on hyperspectral images utilizing the entire spatio-spectral data cube in convolutional neural networks using protein regression analysis of bulk wheat grain kernels as a test case. By introducing novel modifications of the well-known convolutional neural network, ResNet-18, we are able to significantly increase its performance on hyperspectral images. Our modifications consist of firstly applying a 3D convolution layer enabling learning of spectral derivatives that 2D spatial convolution is unable to learn, and secondly, the application of a (1 x 1) 2D convolution layer that downsamples the spectral dimension. Analysis of the responses learned by the convolution kernels in our modifications reveals meaningful representations of the input data cube that reduce noise and enable the subsequent ResNet-18 to perform more accurate regression analysis.
U2 - 10.1109/ICCVW54120.2021.00158
DO - 10.1109/ICCVW54120.2021.00158
M3 - Article in proceedings
SP - 1372
EP - 1380
BT - Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)
PB - IEEE
Y2 - 11 October 2021 through 17 October 2021
ER -
ID: 287119402