Arctic Vision: Using Neural Networks for Ice Object Classification, and Controlling How They Fail
Department of Marine Technology, NTNU Norwegian University of Science and Technology, 7491 Trondheim, Norway
*
Author to whom correspondence should be addressed.
J. Mar. Sci. Eng. 2020, 8(10), 770; https://doi.org/10.3390/jmse8100770
Received: 3 September 2020 / Revised: 25 September 2020 / Accepted: 28 September 2020 / Published: 30 September 2020
(This article belongs to the Special Issue Human-Automation Integration in the Maritime Sector)
Convolutional neural networks (CNNs) have been shown to be excellent at performing image analysis tasks in recent years. Even so, ice object classification using close-range optical images is an area where their use has barely been touched upon, and how well CNNs perform this classification task is still an open question, especially in the challenging visual conditions often found in the High Arctic. The present study explores the use of CNNs for such ice object classification, including analysis of how visual distortion of optical images impacts their performance and comparisons to human experts and novices. To account for the model’s tendency to predict the presence of very few classes for any given image, the use of a loss-weighting scheme pushing a model towards predicting a higher number of classes is proposed. The results of this study show that on clean images, given the class definitions and labeling scheme used, the networks perform better than some humans. At least for some classes of ice objects, the results indicate that the network learned meaningful features. However, the results also indicate that humans are much better at adapting to new visual conditions than neural networks.
View Full-Text
▼
Show Figures
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited
MDPI and ACS Style
Pedersen, O.-M.; Kim, E. Arctic Vision: Using Neural Networks for Ice Object Classification, and Controlling How They Fail. J. Mar. Sci. Eng. 2020, 8, 770. https://doi.org/10.3390/jmse8100770
AMA Style
Pedersen O-M, Kim E. Arctic Vision: Using Neural Networks for Ice Object Classification, and Controlling How They Fail. Journal of Marine Science and Engineering. 2020; 8(10):770. https://doi.org/10.3390/jmse8100770
Chicago/Turabian StylePedersen, Ole-Magnus; Kim, Ekaterina. 2020. "Arctic Vision: Using Neural Networks for Ice Object Classification, and Controlling How They Fail" J. Mar. Sci. Eng. 8, no. 10: 770. https://doi.org/10.3390/jmse8100770
Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.
Search more from Scilit