TY - GEN
T1 - Semantic Segmentation of Underwater Environments Using DeepLabv3+ and Transfer Learning
AU - Chicchon, Miguel
AU - Bedon, Hector
N1 - Funding Information:
Acknowledgements This work has been supported by Programa Nacional de Innovación en Pesca y Acuicultura, PNIPA, Perú, Grant Nr. PNIPA-PES-SIA-PP-000004.
Funding Information:
This work has been supported by Programa Nacional de Innovaci?n en Pesca y Acuicultura, PNIPA, Per?, Grant Nr. PNIPA-PES-SIA-PP-000004.
Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
PY - 2021/10/26
Y1 - 2021/10/26
N2 - The semantic segmentation approach is essential in automated scene analysis, but its application in underwater environments is still limited. Datasets generally have insufficient labeled data, unbalanced data classes, and different lighting conditions, making it difficult to obtain optimal results. Currently, deep convolutional neural networks allow very good results in machine vision tasks, and one of the network architectures with good performance in semantic segmentation is DeepLabv3 +. This paper evaluates the performance of DeepLabv3 + and transfer learning based on pre-trained backend networks in ImageNet to study underwater scenes. The experimentation is carried out on a dataset available on the Internet with labels of eight classes. Experimental results show that DeepLabv3 + and transfer learning are effective for semantic segmentation of multiple underwater scene objects with insufficient tagged data and unbalanced classes.
AB - The semantic segmentation approach is essential in automated scene analysis, but its application in underwater environments is still limited. Datasets generally have insufficient labeled data, unbalanced data classes, and different lighting conditions, making it difficult to obtain optimal results. Currently, deep convolutional neural networks allow very good results in machine vision tasks, and one of the network architectures with good performance in semantic segmentation is DeepLabv3 +. This paper evaluates the performance of DeepLabv3 + and transfer learning based on pre-trained backend networks in ImageNet to study underwater scenes. The experimentation is carried out on a dataset available on the Internet with labels of eight classes. Experimental results show that DeepLabv3 + and transfer learning are effective for semantic segmentation of multiple underwater scene objects with insufficient tagged data and unbalanced classes.
KW - Computer vision
KW - Convolutional neural network
KW - Deep learning
KW - DeepLabv3+
KW - Semantic segmentation
KW - Transfer learning
KW - Underwater images
UR - https://hdl.handle.net/20.500.12724/17607
UR - http://www.scopus.com/inward/record.url?scp=85118997690&partnerID=8YFLogxK
U2 - 10.1007/978-981-16-4016-2_29
DO - 10.1007/978-981-16-4016-2_29
M3 - Articulo (Contribución a conferencia)
AN - SCOPUS:85118997690
SN - 9789811640155
VL - 286
T3 - Lecture Notes in Networks and Systems
SP - 301
EP - 309
BT - Smart Trends in Computing and Communications - Proceedings of SmartCom 2021
A2 - Zhang, Yu-Dong
A2 - Senjyu, Tomonobu
A2 - So-In, Chakchai
A2 - Joshi, Amit
PB - Springer Nature
T2 - 5th International Conference on Smart Trends in Computing and Communications, SmartCom 2021
Y2 - 15 April 2021 through 16 April 2021
ER -