{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,5]],"date-time":"2026-03-05T14:05:54Z","timestamp":1772719554233,"version":"3.50.1"},"reference-count":41,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,1,8]],"date-time":"2025-01-08T00:00:00Z","timestamp":1736294400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100022591","name":"Energiteknologisk udviklings- og demonstrationsprogram","doi-asserted-by":"publisher","award":["64020-1093"],"award-info":[{"award-number":["64020-1093"]}],"id":[{"id":"10.13039\/501100022591","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Robot. AI"],"abstract":"<jats:sec><jats:title>Introduction<\/jats:title><jats:p>Subsea applications recently received increasing attention due to the global expansion of offshore energy, seabed infrastructure, and maritime activities; complex inspection, maintenance, and repair tasks in this domain are regularly solved with pilot-controlled, tethered remote-operated vehicles to reduce the use of human divers. However, collecting and precisely labeling submerged data is challenging due to uncontrollable and harsh environmental factors. As an alternative, synthetic environments offer cost-effective, controlled alternatives to real-world operations, with access to detailed ground-truth data. This study investigates the potential of synthetic underwater environments to offer cost-effective, controlled alternatives to real-world operations, by rendering detailed labeled datasets and their application to machine-learning.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>Two synthetic datasets with over 1000 rendered images each were used to train DeepLabV3+ neural networks with an Xception backbone. The dataset includes environmental classes like seawater and seafloor, offshore structures components, ship hulls, and several marine growth classes. The machine-learning models were trained using transfer learning and data augmentation techniques.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Testing showed high accuracy in segmenting synthetic images. In contrast, testing on real-world imagery yielded promising results for two out of three of the studied cases, though challenges in distinguishing some classes persist.<\/jats:p><\/jats:sec><jats:sec><jats:title>Discussion<\/jats:title><jats:p>This study demonstrates the efficiency of synthetic environments for training subsea machine learning models but also highlights some important limitations in certain cases. Improvements can be pursued by introducing layered species into synthetic environments and improving real-world optical information quality\u2014better color representation, reduced compression artifacts, and minimized motion blur\u2014are key focus areas. Future work involves more extensive validation with expert-labeled datasets to validate and enhance real-world application accuracy.<\/jats:p><\/jats:sec>","DOI":"10.3389\/frobt.2024.1459570","type":"journal-article","created":{"date-parts":[[2025,1,8]],"date-time":"2025-01-08T11:11:41Z","timestamp":1736334701000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Semantic segmentation using synthetic images of underwater marine-growth"],"prefix":"10.3389","volume":"11","author":[{"given":"Christian","family":"Mai","sequence":"first","affiliation":[]},{"given":"Jesper","family":"Liniger","sequence":"additional","affiliation":[]},{"given":"Simon","family":"Pedersen","sequence":"additional","affiliation":[]}],"member":"1965","published-online":{"date-parts":[[2025,1,8]]},"reference":[{"key":"B2","doi-asserted-by":"publisher","first-page":"1456","DOI":"10.1002\/rob.21915","article-title":"CoralSeg: learning coral segmentation from sparse annotations","volume":"36","author":"Alonso","year":"2019","journal-title":"J. Field Robotics"},{"key":"B3","unstructured":"Synthetic image datasets\n          \n          \n          2023"},{"key":"B1","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/OCEANSLimerick52467.2023.10244578","article-title":"Multi-label video classification for underwater ship inspection","volume-title":"Oceans 2023 - limerick","author":"Azad","year":"2023"},{"key":"B4","doi-asserted-by":"publisher","first-page":"437","DOI":"10.1007\/978-3-642-35289-8_26","article-title":"Practical recommendations for gradient-based training of deep architectures","author":"Bengio","year":"2012","journal-title":"Lect. Notes Comput. Sci. Incl. Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinforma. 7700 Lect. NO"},{"key":"B5","unstructured":"Rendering - Cycles - introduction \u2014 blender manual\n          \n          \n          2023"},{"key":"B6","doi-asserted-by":"publisher","first-page":"2095","DOI":"10.3390\/f13122095","article-title":"Image segmentation method for sweetgum leaf spots based on an improved DeeplabV3+ network","volume":"12","author":"Cai","year":"2022","journal-title":"For. 13"},{"key":"B7","doi-asserted-by":"publisher","first-page":"833","DOI":"10.1007\/978-3-030-01234-2_49","article-title":"Encoder-Decoder with atrous separable convolution for semantic image segmentation","author":"Chen","year":"2018","journal-title":"Lect. Notes Comput. Sci. Incl. Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinforma."},{"key":"B8","article-title":"Intelligent image recognition system for marine fouling using softmax transfer learning and deep convolutional neural networks","volume-title":"Complexity","author":"Chin","year":"2017"},{"key":"B9","doi-asserted-by":"publisher","first-page":"1800","DOI":"10.1109\/CVPR.2017.195","article-title":"Xception: deep learning with depthwise separable convolutions","author":"Chollet","year":"2017","journal-title":"2017 IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR)"},{"key":"B10","doi-asserted-by":"publisher","article-title":"North West Shelf Region, Bio-Geo-Chemical, L4, monthly means and interpolated daily observation","year":"2023","DOI":"10.48670\/moi-00119"},{"key":"B11","article-title":"Marine growth effects on the hydrodynamical behaviour of circular structures","volume-title":"14\u00e8mes journ\u00e9es de l\u2019Hydrodynamique","author":"Gaurier","year":"2014"},{"key":"B12","first-page":"1","article-title":"The importance of using effective anti-fouling coatings in relation to greenhouse gas emissions from shipping","year":"2010","journal-title":"Submitt. by Int. Paint Print. Ink Coucil (IPPIC)"},{"key":"B13","doi-asserted-by":"crossref","first-page":"1769","DOI":"10.1109\/IROS45743.2020.9340821","article-title":"Semantic segmentation of underwater imagery: dataset and benchmark","volume-title":"2020 IEEE\/RSJ international conference on intelligent robots and systems (IROS)","author":"Islam","year":"2020"},{"key":"B14","volume-title":"Artificial intelligence for automated marine growth classification","year":"2023"},{"key":"B15","doi-asserted-by":"publisher","first-page":"108467","DOI":"10.1016\/J.ENGFRACMECH.2022.108467","article-title":"Automatic surface crack detection using segmentation-based deep-learning approach","volume":"268","author":"Joshi","year":"2022","journal-title":"Eng. Fract. Mech."},{"key":"B16","first-page":"279","article-title":"Semantic scene segmentation for robotics","volume-title":"Deep learning for robot perception and cognition","year":"2024"},{"key":"B17","first-page":"1","article-title":"DatUS: data-driven unsupervised semantic segmentation with pre-trained self-supervised vision transformer","volume-title":"IEEE transactions on cognitive and developmental systems","author":"Kumar","year":"2024"},{"key":"B18","doi-asserted-by":"crossref","first-page":"255","DOI":"10.1109\/WACVW58289.2023.00032","article-title":"A survey on the deployability of semantic segmentation networks for fluvial navigation","volume-title":"Proceedings - 2023 IEEE\/CVF winter conference on applications of computer vision workshops, WACVW 2023","author":"Lambert","year":"2023"},{"key":"B19","doi-asserted-by":"crossref","DOI":"10.1109\/OCEANS47191.2022.9977233","article-title":"OysterSim: underwater simulation for enhancing oyster reef monitoring","volume-title":"Oceans conference record (IEEE)","author":"Lin","year":"2022"},{"key":"B20","first-page":"1","article-title":"On the autonomous inspection and classification of marine growth on subsea structures","volume-title":"Proceedings of the IEEE OCEANS 2022 conference","author":"Liniger","year":"2022"},{"key":"B21","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1016\/j.undsp.2023.09.012","article-title":"Transfer learning-based encoder-decoder model with visual explanations for infrastructure crack segmentation: New open database and comprehensive evaluation","volume":"17","author":"Liu","year":"2024","journal-title":"Undergr. Space"},{"key":"B22","doi-asserted-by":"publisher","first-page":"102531","DOI":"10.1016\/j.jag.2021.102531","article-title":"Study on transfer learning ability for classifying marsh vegetation with multi-sensor images using DeepLabV3+ and HRNet deep learning algorithms","volume":"103","author":"Liu","year":"2021","journal-title":"Int. J. Appl. Earth Observation Geoinformation"},{"key":"B23","doi-asserted-by":"publisher","first-page":"64","DOI":"10.1080\/08927014.2023.2185143","article-title":"Semantic segmentation for fully automated macrofouling analysis on coatings after field exposure","volume":"39","author":"LutzKrause","year":"2023","journal-title":"Biofouling"},{"key":"B24","doi-asserted-by":"publisher","first-page":"64119284","DOI":"10.1016\/j.oceaneng.2024.119284","article-title":"Synthetic subsea imagery for marine-growth inspection under natural lighting","volume":"313","author":"Mai","year":"2024","journal-title":"Ocean Eng."},{"key":"B25","unstructured":"Real-time semantic segmentation using hyperspectral images for mapping unstructured and unknown environments\n          \n          \n            \n              Medellin\n              A.\n            \n            \n              Bhamri\n              A.\n            \n            \n              Langari\n              R.\n            \n            \n              Gopalswamy\n              S.\n            \n          \n          \n          2023"},{"key":"B26","doi-asserted-by":"publisher","first-page":"982581","DOI":"10.3389\/frobt.2022.982581","article-title":"Framework for environment perception: ensemble method for vision-based scene understanding algorithms in agriculture","volume":"9","author":"Mujkic","year":"2023","journal-title":"Front. Robotics AI"},{"key":"B27","doi-asserted-by":"publisher","first-page":"374","DOI":"10.20517\/IR.2023.22","article-title":"Deep learning-based scene understanding for autonomous robots: a survey","author":"Ni","year":"2023","journal-title":"Intell. Robotics"},{"key":"B28","doi-asserted-by":"publisher","first-page":"012130","DOI":"10.1088\/1742-6596\/628\/1\/012130","article-title":"Protocols for image processing based underwater inspection of infrastructure elements","volume":"628","author":"O\u2019Byrne","year":"2015","journal-title":"J. Phys. Conf. Ser."},{"key":"B29","doi-asserted-by":"publisher","first-page":"93","DOI":"10.3390\/jmse6030093","article-title":"Semantic segmentation of underwater imagery using deep networks trained on synthetic imagery","volume":"6","author":"O\u2019Byrne","year":"","journal-title":"J. Mar. Sci. Eng."},{"key":"B30","doi-asserted-by":"publisher","first-page":"104","DOI":"10.1080\/15732479.2017.1330890","article-title":"An underwater lighting and turbidity image repository for analysing the performance of image-based non-destructive techniques","volume":"14","author":"O\u2019Byrne","year":"","journal-title":"Struct. Infrastructure Eng."},{"key":"B31","doi-asserted-by":"publisher","first-page":"195","DOI":"10.1007\/978-1-4684-7697-2","article-title":"Computer processing of remotely-sensed images","author":"Paul","year":"2012"},{"key":"B32","first-page":"1","article-title":"On marine growth removal on offshore structures","volume-title":"Oceans conference record (IEEE)","author":"Pedersen","year":"2022"},{"key":"B33","doi-asserted-by":"publisher","first-page":"2144","DOI":"10.3390\/S21062144","article-title":"BLAINDER\u2014a blender AI add-on for generation of semantically labeled depth-sensing data","volume":"21","author":"Reitmann","year":"2021","journal-title":"Sensors 2021"},{"key":"B34","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-75178-4","article-title":"Nikolenko. Synthetic data for deep learning","volume":"174","author":"Sergey","year":"2021","journal-title":"Springer Optim. Its Appl."},{"key":"B35","doi-asserted-by":"publisher","first-page":"101085","DOI":"10.1016\/J.ECOINF.2020.101085","article-title":"\u201cHow many images do I need?\u201d Understanding how sample size per class affects deep learning model performance metrics for balanced designs in autonomous wildlife monitoring","volume":"57","author":"Shahinfar","year":"2020","journal-title":"Ecol. Inf."},{"key":"B36","doi-asserted-by":"crossref","DOI":"10.1109\/OCEANS47191.2022.9977212","article-title":"A deep learning framework for semantic segmentation of underwater environments","volume-title":"Oceans conference record (IEEE)","author":"Smith","year":"2022"},{"key":"B37","unstructured":"Inconsistency masks: removing the uncertainty from input-pseudo-label pairs\n          \n          \n            \n              Vorndran\n              M. R. H.\n            \n            \n              Roeck\n              B. F.\n            \n          \n          \n          2024"},{"key":"B38","doi-asserted-by":"publisher","first-page":"462","DOI":"10.1109\/JOE.2022.3219129","article-title":"Semantic segmentation in underwater ship inspections: benchmark and data set","volume":"48","author":"Waszak","year":"2023","journal-title":"IEEE J. Ocean. Eng."},{"key":"B39","first-page":"997","article-title":"Physically based sonar simulation and image generation for side-scan sonar","volume-title":"Uace 2017","author":"Woock","year":"2017"},{"key":"B40","first-page":"3","article-title":"AquaSAM: underwater image foreground segmentation","volume-title":"Communications in Computer and information science","author":"Xu","year":"2024"},{"key":"B41","doi-asserted-by":"publisher","first-page":"204","DOI":"10.1016\/J.NEUCOM.2023.01.056","article-title":"A systematic review and analysis of deep learning-based underwater object detection","volume":"527","author":"Xu","year":"2023","journal-title":"Neurocomputing"}],"container-title":["Frontiers in Robotics and AI"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frobt.2024.1459570\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,1,8]],"date-time":"2025-01-08T11:11:50Z","timestamp":1736334710000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frobt.2024.1459570\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,1,8]]},"references-count":41,"alternative-id":["10.3389\/frobt.2024.1459570"],"URL":"https:\/\/doi.org\/10.3389\/frobt.2024.1459570","relation":{},"ISSN":["2296-9144"],"issn-type":[{"value":"2296-9144","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,1,8]]},"article-number":"1459570"}}