{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,28]],"date-time":"2026-04-28T05:04:53Z","timestamp":1777352693954,"version":"3.51.4"},"reference-count":48,"publisher":"MDPI AG","issue":"6","license":[{"start":{"date-parts":[[2017,6,2]],"date-time":"2017-06-02T00:00:00Z","timestamp":1496361600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Spanish Economics and Competitiveness Ministry","award":["AYA2014-57648-P"],"award-info":[{"award-number":["AYA2014-57648-P"]}]},{"name":"Government of the Principality of Asturias (Consejer\u00eda de Econom\u00eda y Empleo)","award":["FC-15-GRUPIN14-017"],"award-info":[{"award-number":["FC-15-GRUPIN14-017"]}]},{"name":"UK Science and Technology Facilities Council","award":["ST\/K003569\/1"],"award-info":[{"award-number":["ST\/K003569\/1"]}]},{"name":"UK Science and Technology Facilities Council","award":["ST\/L00075X\/1"],"award-info":[{"award-number":["ST\/L00075X\/1"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>Many of the next generation of adaptive optics systems on large and extremely large telescopes require tomographic techniques in order to correct for atmospheric turbulence over a large field of view. Multi-object adaptive optics is one such technique. In this paper, different implementations of a tomographic reconstructor based on a machine learning architecture named \u201cCARMEN\u201d are presented. Basic concepts of adaptive optics are introduced first, with a short explanation of three different control systems used on real telescopes and the sensors utilised. The operation of the reconstructor, along with the three neural network frameworks used, and the developed CUDA code are detailed. Changes to the size of the reconstructor influence the training and execution time of the neural network. The native CUDA code turns out to be the best choice for all the systems, although some of the other frameworks offer good performance under certain circumstances.<\/jats:p>","DOI":"10.3390\/s17061263","type":"journal-article","created":{"date-parts":[[2017,6,2]],"date-time":"2017-06-02T10:20:44Z","timestamp":1496398844000},"page":"1263","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":17,"title":["Comparative Study of Neural Network Frameworks for the Next Generation of Adaptive Optics Systems"],"prefix":"10.3390","volume":"17","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9912-7807","authenticated-orcid":false,"given":"Carlos","family":"Gonz\u00e1lez-Guti\u00e9rrez","sequence":"first","affiliation":[{"name":"Mining Exploitation and Prospecting Department, University of Oviedo, 33004 Oviedo, Spain"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2038-4606","authenticated-orcid":false,"given":"Jes\u00fas","family":"Santos","sequence":"additional","affiliation":[{"name":"Department of Physics, University of Oviedo, 33004 Oviedo, Spain"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6866-3316","authenticated-orcid":false,"given":"Mario","family":"Mart\u00ednez-Zarzuela","sequence":"additional","affiliation":[{"name":"Department of Signal Theory and Communications and Telematics Engineering, University of Valladolid, 47011 Valladolid, Spain"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0904-7097","authenticated-orcid":false,"given":"Alistair","family":"Basden","sequence":"additional","affiliation":[{"name":"Department of Physics, Centre for Advanced Instrumentation, University of Durham, South Road, Durham DH1 3LE, UK"}]},{"given":"James","family":"Osborn","sequence":"additional","affiliation":[{"name":"Department of Physics, Centre for Advanced Instrumentation, University of Durham, South Road, Durham DH1 3LE, UK"}]},{"given":"Francisco","family":"D\u00edaz-Pernas","sequence":"additional","affiliation":[{"name":"Department of Signal Theory and Communications and Telematics Engineering, University of Valladolid, 47011 Valladolid, Spain"}]},{"given":"Francisco","family":"De Cos Juez","sequence":"additional","affiliation":[{"name":"Mining Exploitation and Prospecting Department, University of Oviedo, 33004 Oviedo, Spain"}]}],"member":"1968","published-online":{"date-parts":[[2017,6,2]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"21356","DOI":"10.1364\/OE.18.021356","article-title":"Modeling a MEMS deformable mirror using non-parametric estimation techniques","volume":"18","author":"Juez","year":"2010","journal-title":"Opt. Express"},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"6354","DOI":"10.1364\/AO.49.006354","article-title":"Durham adaptive optics real-time controller","volume":"49","author":"Basden","year":"2010","journal-title":"Appl. Opt."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"6492","DOI":"10.1364\/OE.18.006492","article-title":"Deformable mirror model for open-loop adaptive optics using multivariate adaptive regression splines","volume":"18","author":"Lasheras","year":"2010","journal-title":"Opt. Express"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"783","DOI":"10.1364\/JOSAA.11.000783","article-title":"First-order performance evaluation of adaptive-optics systems for atmospheric-turbulence compensation in extended-field-of-view astronomical telescopes","volume":"11","author":"Ellerbroek","year":"1994","journal-title":"J. Opt. Soc. Am. A"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"2527","DOI":"10.1364\/JOSAA.18.002527","article-title":"Optimal wave-front reconstruction strategies for multiconjugate adaptive optics","volume":"18","author":"Fusco","year":"2001","journal-title":"J. Opt. Soc. Am. A"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"451","DOI":"10.1016\/0045-7906(92)90038-F","article-title":"Optical performance of fully and partially compensated adaptive optics systems using least-squares and minimum variance phase reconstructors","volume":"18","author":"Roggemann","year":"1992","journal-title":"Comput. Electr. Eng."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"A253","DOI":"10.1364\/JOSAA.27.00A253","article-title":"Tomography approach for multi-object adaptive optics","volume":"27","author":"Vidal","year":"2010","journal-title":"J. Opt. Soc. Am. A"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"8895","DOI":"10.3390\/s120708895","article-title":"An ANN-based smart tomographic reconstructor in a dynamic environment","volume":"12","author":"Osborn","year":"2012","journal-title":"Sensors"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"20382","DOI":"10.3390\/s141120382","article-title":"Missing Data Imputation of Solar Radiation Data under Different Atmospheric Conditions","volume":"14","author":"Turrado","year":"2014","journal-title":"Sensors"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"469","DOI":"10.1086\/665924","article-title":"Performance Modeling for the RAVEN Multi-Object Adaptive Optics Demonstrator","volume":"124","author":"Andersen","year":"2012","journal-title":"Publ. Astron. Soc. Pac."},{"key":"ref_11","doi-asserted-by":"crossref","unstructured":"Lardi\u00e8re, O., Andersen, D., Blain, C., Bradley, C., Gamroth, D., Jackson, K., Lach, P., Nash, R., Venn, K., and V\u00e9ran, J.-P. (2014, January 22). Multi-object adaptive optics on-sky results with Raven. Proceedings of the SPIE 9148, Adaptive Optics Systems IV, Montreal, QC, Canada.","DOI":"10.1117\/12.2055480"},{"key":"ref_12","first-page":"351","article-title":"A Hybrid Device of Self Organizing Maps (SOM) and Multivariate Adaptive Regression Splines (MARS) for the Forecasting of Firms\u2019","volume":"10","author":"Lorca","year":"2011","journal-title":"J. Account. Manag. Inf. Syst."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"68","DOI":"10.1016\/j.ecoleng.2012.12.015","article-title":"Forecasting the cyanotoxins presence in fresh waters: A new model based on genetic algorithms combined with the MARS technique","volume":"53","author":"Niz","year":"2013","journal-title":"Ecol. Eng."},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Su\u00e1rez G\u00f3mez, S., Santos Rodr\u00edguez, J., Iglesias Rodr\u00edguez, F., and de Cos Juez, F. (2017). Analysis of the Temporal Structure Evolution of Physical Systems with the Self-Organising Tree Algorithm (SOTA): Application for Validating Neural Network Systems on Adaptive Optics Data before On-Sky Implementation. Entropy, 19.","DOI":"10.3390\/e19030103"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"3457","DOI":"10.1007\/s11269-013-0358-4","article-title":"Support Vector Machines and Multilayer Perceptron Networks Used to Evaluate the Cyanotoxins Presence from Experimental Cyanobacteria Concentrations in the Trasona Reservoir (Northern Spain)","volume":"27","year":"2013","journal-title":"Water Resour. Manag."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Casteleiro-Roca, J.-L., Calvo-Rolle, J., M\u00e9ndez P\u00e9rez, J., Roque\u00f1\u00ed Guti\u00e9rrez, N., and de Cos Juez, F. (2017). Hybrid Intelligent System to Perform Fault Detection on BIS Sensor During Surgeries. Sensors, 17.","DOI":"10.3390\/s17010179"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"92","DOI":"10.1016\/j.ergon.2015.09.012","article-title":"Applying the K-nearest neighbor technique to the classification of workers according to their risk of suffering musculoskeletal disorders","volume":"52","year":"2016","journal-title":"Int. J. Ind. Ergon."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"952","DOI":"10.1016\/j.amc.2008.10.019","article-title":"Non-linear numerical analysis of a double-threaded titanium alloy dental implant by FEM","volume":"206","year":"2008","journal-title":"Appl. Math. Comput."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"1665","DOI":"10.1016\/j.mcm.2010.11.069","article-title":"Application of neural networks to the study of the influence of diet and lifestyle on the value of bone mineral density in post-menopausal women","volume":"54","year":"2011","journal-title":"Math. Comput. Model."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"37","DOI":"10.1016\/j.resourpol.2015.03.004","article-title":"Forecasting the COMEX copper spot price by means of neural networks and ARIMA models","volume":"45","author":"Lasheras","year":"2015","journal-title":"Resour. Policy"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"2420","DOI":"10.1364\/OE.20.002420","article-title":"Using artificial neural networks for open-loop tomography","volume":"20","author":"Osborn","year":"2012","journal-title":"Opt. Express"},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"2508","DOI":"10.1093\/mnras\/stu758","article-title":"Open-loop tomography with artificial neural networks on CANARY: On-sky results","volume":"441","author":"Osborn","year":"2014","journal-title":"Mon. Not. R. Astron. Soc."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Marchetti, E., Close, L.M., and V\u00e9ran, J.-P. (2014). First on-sky results of a neural network based tomographic reconstructor: CARMEN on CANARY. SPIE Astronomical Telescopes + Instrumentation, International Society for Optics and Photonics.","DOI":"10.1117\/12.2057462"},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Ramsay, S.K., Casali, M.M., Gonz\u00e1lez, J.C., and Hubin, N. (2014, January 22). The E-ELT instrument roadmap: A status report. Proceedings of the SPIE 9147, Ground-based and Airborne Instrumentation for Astronomy V, Montreal, QC, Canada.","DOI":"10.1117\/12.2056341"},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"7587","DOI":"10.1364\/AO.44.007587","article-title":"Atmospheric wavefront phase recovery by use of specialized hardware: Graphical processing units and field-programmable gate arrays","volume":"44","author":"Rosa","year":"2005","journal-title":"Appl. Opt."},{"key":"ref_26","unstructured":"Ltaief, H., and Gratadour, D. (2016, March 14). Shooting for the Stars with GPUs. Available online: http:\/\/on-demand.gputechconf.com\/gtc\/2015\/video\/S5122.html."},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Gonz\u00e1lez-Guti\u00e9rrez, C., Santos-Rodr\u00edguez, J.D., D\u00edaz, R.\u00c1.F., Rolle, J.L.C., Guti\u00e9rrez, N.R., and de Cos Juez, F.J. (2017). Using GPUs to Speed up a Tomographic Reconstructor Based on Machine Learning. International Joint Conference SOCO\u201916-CISIS\u201916-ICEUTE\u201916, Springer.","DOI":"10.1007\/978-3-319-47364-2_27"},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Su\u00e1rez G\u00f3mez, S.L., Gonz\u00e1lez-Guti\u00e9rrez, C., Santos-Rodr\u00edguez, J.D., S\u00e1nchez Rodr\u00edguez, M.L., S\u00e1nchez Lasheras, F., and de Cos Juez, F.J. (2016, January 14\u201316). Analysing the performance of a tomographic reconstructor with different neural networks frameworks. Proceedings of the International Conference on Intelligent Systems Design and Applications, Porto, Portugal.","DOI":"10.1007\/978-3-319-53480-0_103"},{"key":"ref_29","unstructured":"Gulcehre, C. (2017, February 15). Deep Learning\u2014Software Links. Available online: http:\/\/deeplearning.net\/software_links\/."},{"key":"ref_30","unstructured":"Bahrampour, S., Ramakrishnan, N., Schott, L., and Shah, M. (2015). Comparative Study of Deep Learning Software Frameworks. arXiv."},{"key":"ref_31","unstructured":"Soumith, C. (2016, May 20). Convnet-Benchmarks. Available online: https:\/\/github.com\/soumith\/convnet-benchmarks."},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Shi, S., Wang, Q., Xu, P., and Chu, X. (2016). Benchmarking State-of-the-Art Deep Learning Software Tools. arXiv.","DOI":"10.1109\/CCBD.2016.029"},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"S573","DOI":"10.3928\/1081-597X-20010901-13","article-title":"History and Principles of Shack-Hartmann Wavefront Sensing","volume":"17","author":"Platt","year":"2001","journal-title":"J. Refract. Surg."},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"998","DOI":"10.1364\/JOSA.70.000998","article-title":"Wave-front estimation from wave-front slope measurements","volume":"70","author":"Southwell","year":"1980","journal-title":"J. Opt. Soc. Am."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"1350","DOI":"10.1093\/mnras\/stw730","article-title":"Experience with wavefront sensor and deformable mirror interfaces for wide-field adaptive optics systems","volume":"459","author":"Basden","year":"2016","journal-title":"Mon. Not. R. Astron. Soc"},{"key":"ref_36","unstructured":"Morris, T., Gendron, E., Basden, A., Martin, O., Osborn, J., Henry, D., Hubert, Z., Sivo, G., Gratadour, D., and Chemla, F. (2013, January 26\u201331). Multiple Object Adaptive Optics: Mixed NGS\/LGS tomography. Proceedings of the Third AO4ELT Conference, Firenze, Italy."},{"key":"ref_37","unstructured":"Chanan, G. (2000). Principles of Wavefront Sensing and Reconstruction. Proceedings: Summer School on Adaptive Optics, Center for Adaptive Optics (CfAO)."},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"53101","DOI":"10.1117\/1.OE.53.5.053101","article-title":"Single exposure compressed imaging system with Hartmann-Shack wavefront sensor","volume":"53","author":"Xiao","year":"2014","journal-title":"Opt. Eng."},{"key":"ref_39","doi-asserted-by":"crossref","unstructured":"Myers, R.M., Hubert, Z., Morris, T.J., Gendron, E., Dipper, N.A., Kellerer, A., Goodsell, S.J., Rousset, G., Younger, E., and Marteaud, M. (2008). CANARY: The on-sky NGS\/LGS MOAO demonstrator for EAGLE. SPIE Astronomical Telescopes+ Instrumentation, International Society for Optics and Photonics.","DOI":"10.1117\/12.789544"},{"key":"ref_40","doi-asserted-by":"crossref","unstructured":"Marchetti, E., Close, L.M., and V\u00e9ran, J.-P. (2014). DRAGON, the Durham real-time, tomographic adaptive optics test bench: Progress and results. SPIE Astronomical Telescopes+ Instrumentation, International Society for Optics and Photonics.","DOI":"10.1117\/12.2055415"},{"key":"ref_41","unstructured":"Dipper, N.A., Basden, A., Bitenc, U., Myers, R.M., Richards, A., and Younger, E.J. (2013, January 26\u201331). Adaptive Optics for Extremely Large Telescopes III ADAPTIVE OPTICS REAL-TIME CONTROL SYSTEMS FOR THE E-ELT. Proceedings of the Adaptive Optics for Extremely Large Telescopes III, Florence, Italy."},{"key":"ref_42","doi-asserted-by":"crossref","unstructured":"Osborn, J., De Cos Juez, F.J., Guzman, D., Butterley, T., Myers, R., Guesalaga, A., and Laine, J. (2011, January 25\u201330). Open-loop tomography using artificial nueral networks. Proceedings of the Adaptive Optics for Extremely Large Telescopes II, Victoria, BC, Canada.","DOI":"10.1364\/OE.20.002420"},{"key":"ref_43","unstructured":"G\u00f3mez Victoria, M. (2014). Research of the Tomographic Reconstruction Problem by Means of Data Mining and Artificial Intelligence Technologies. [Ph.D. Thesis, Universidad de Oviedo]."},{"key":"ref_44","unstructured":"Jia, Y., Shelhamer, E., Donahue, J., Karayev, S., Long, J., Girshick, R., Guadarrama, S., and Darrell, T. Caffe: Convolutional Architecture for Fast Feature Embedding. Proceedings of the ACM International Conference on Multimedia\u2014MM \u201914."},{"key":"ref_45","unstructured":"Al-Rfou, R., Alain, G., Almahairi, A., Angermueller, C., Bahdanau, D., Ballas, N., Bastien, F., Bayer, J., Belikov, A., and Belopolsky, A. (2016). Theano: A Python framework for fast computation of mathematical expressions. arXiv."},{"key":"ref_46","unstructured":"(2016, June 20). The HDF Group Introduction to HDF5. Available online: https:\/\/www.hdfgroup.org\/HDF5\/doc\/H5.intro.html#Intro-WhatIs."},{"key":"ref_47","unstructured":"Pereira, F., Burges, C.J.C., Bottou, L., and Weinberger, K.Q. (2012). ImageNet Classification with Deep Convolutional Neural Networks. Advances in Neural Information Processing Systems 25, Curran Associates, Inc."},{"key":"ref_48","doi-asserted-by":"crossref","first-page":"801","DOI":"10.1016\/S0893-6080(05)80125-X","article-title":"Approximation of dynamical systems by continuous time recurrent neural networks","volume":"6","author":"Funahashi","year":"1993","journal-title":"Neural Netw."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/17\/6\/1263\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T18:37:42Z","timestamp":1760207862000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/17\/6\/1263"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2017,6,2]]},"references-count":48,"journal-issue":{"issue":"6","published-online":{"date-parts":[[2017,6]]}},"alternative-id":["s17061263"],"URL":"https:\/\/doi.org\/10.3390\/s17061263","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2017,6,2]]}}}