{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,2]],"date-time":"2026-07-02T02:04:52Z","timestamp":1782957892282,"version":"3.54.5"},"reference-count":63,"publisher":"Acoustical Society of America (ASA)","issue":"4","license":[{"start":{"date-parts":[[2023,4,21]],"date-time":"2023-04-21T00:00:00Z","timestamp":1682035200000},"content-version":"vor","delay-in-days":20,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2023,4,21]],"date-time":"2023-04-21T00:00:00Z","timestamp":1682035200000},"content-version":"tdm","delay-in-days":20,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["pubs.aip.org"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2023,4,1]]},"abstract":"<jats:p>Today's low-cost automotive ultrasonic sensors perform distance measurements of obstacles within the close range of vehicles. For future parking assist systems and autonomous driving applications, the performance of the sensors should be further increased. This paper examines the processing of sensor data for the classification of different object classes and traversability of obstacles using a single ultrasonic sensor. The acquisition of raw time signals, transformation into time-frequency images, and classification using machine learning methods are described. Stationary and dynamic measurements at a velocity of 0.5\u2009m\/s of various objects have been carried out in a semi-anechoic chamber and on an asphalt parking space. We propose a scalogram-based signal processing chain and a convolutional neural network, which outperforms a LeNet-5-like baseline. Additionally, several methods for offline and online data augmentation are presented and evaluated. It is shown that carefully selected augmentation methods are useful to train more robust models. Accuracies of 90.1% are achieved for the classification of seven object classes in the laboratory and 66.4% in the outdoor environment. Traversability is correctly classified at an accuracy of 96.4% and 91.5%, respectively.<\/jats:p>","DOI":"10.1121\/10.0017922","type":"journal-article","created":{"date-parts":[[2023,4,27]],"date-time":"2023-04-27T17:14:23Z","timestamp":1682615663000},"page":"2447","update-policy":"https:\/\/doi.org\/10.1063\/aip-crossmark-policy-page","source":"Crossref","is-referenced-by-count":14,"title":["Convolutional neural network with data augmentation for object classification in automotive ultrasonic sensing"],"prefix":"10.1121","volume":"153","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-7686-0005","authenticated-orcid":false,"given":"Jona","family":"Eisele","sequence":"first","affiliation":[{"name":"Corporate Research (CR), Robert Bosch GmbH 1 , Robert-Bosch-Campus 1, Renningen 71272, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0816-7177","authenticated-orcid":false,"given":"Andr\u00e9","family":"Gerlach","sequence":"additional","affiliation":[{"name":"Corporate Research (CR), Robert Bosch GmbH 1 , Robert-Bosch-Campus 1, Renningen 71272, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8541-8630","authenticated-orcid":false,"given":"Marcus","family":"Maeder","sequence":"additional","affiliation":[{"name":"Chair of Vibroacoustics of Vehicles and Machines, Technical University of Munich 2 , Boltzmannstrasse 15, Garching near Munich 85748, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6464-8015","authenticated-orcid":false,"given":"Steffen","family":"Marburg","sequence":"additional","affiliation":[{"name":"Chair of Vibroacoustics of Vehicles and Machines, Technical University of Munich 2 , Boltzmannstrasse 15, Garching near Munich 85748, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"231","published-online":{"date-parts":[[2023,4,21]]},"reference":[{"key":"2023081107265211100_c1a","doi-asserted-by":"crossref","first-page":"303","DOI":"10.1007\/978-3-319-12352-3","article-title":"Ultrasonic sensors for a K44DAS","volume-title":"Handbook of Driver Assistance Systems","author":"Winner","year":"2016"},{"key":"2023081107265211100_c2","doi-asserted-by":"publisher","first-page":"261","DOI":"10.1016\/j.eng.2018.11.020","article-title":"Deep learning in medical ultrasound analysis: A review","volume":"5","year":"2019","journal-title":"Engineering"},{"key":"2023081107265211100_c3","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1016\/j.media.2017.07.005","article-title":"A survey on deep learning in medical image analysis","volume":"42","year":"2017","journal-title":"Med. Image Anal."},{"key":"2023081107265211100_c4","doi-asserted-by":"publisher","first-page":"87","DOI":"10.1016\/0963-8695(95)00053-4","article-title":"Neural network classification of flaws detected by ultrasonic means","volume":"29","year":"1996","journal-title":"NDT&E Int."},{"key":"2023081107265211100_c5","doi-asserted-by":"publisher","first-page":"20","DOI":"10.1007\/s10921-010-0086-0","article-title":"Automatic defect classification in ultrasonic NDT using artificial intelligence","volume":"30","year":"2011","journal-title":"J. Nondestruct. Eval."},{"key":"2023081107265211100_c6","doi-asserted-by":"publisher","first-page":"1972","DOI":"10.3390\/electronics9111972","article-title":"A review on deep learning-based approaches for automatic sonar target recognition","volume":"9","year":"2020","journal-title":"Electronics"},{"key":"2023081107265211100_c7","doi-asserted-by":"crossref","first-page":"637","DOI":"10.1007\/978-3-540-69460-1","article-title":"Ultrasound","volume-title":"Handbook of Engineering Acoustics","author":"M\u00fcller","year":"2013"},{"key":"2023081107265211100_c8","doi-asserted-by":"crossref","first-page":"407","DOI":"10.1007\/978-1-4615-0312-5_7","article-title":"The spatial representation","volume-title":"Echo Signal Processing","year":"2003"},{"key":"2023081107265211100_c9","first-page":"1","article-title":"Terrain classification for autonomous vehicles using bat-inspired echolocation","year":"2018"},{"key":"2023081107265211100_c10","doi-asserted-by":"publisher","first-page":"19","DOI":"10.1016\/j.proeng.2016.11.119","article-title":"Road surface classification using automotive ultrasonic sensor","volume":"168","year":"2016","journal-title":"Procedia Eng."},{"key":"2023081107265211100_c11","first-page":"661","article-title":"Capsule neural network based height classification using low-cost automotive ultrasonic sensors","year":"2019"},{"key":"2023081107265211100_c12","first-page":"1","article-title":"Ultrasonic sensor-based human detector using one-class classifiers","year":"2015"},{"key":"2023081107265211100_c13","first-page":"672","article-title":"Stair case detection and recognition using ultrasonic signal","year":"2013"},{"key":"2023081107265211100_c14","doi-asserted-by":"publisher","first-page":"141","DOI":"10.1016\/0003-3472(60)90022-1","article-title":"The echolocation of flying insects by bats","volume":"8","year":"1960","journal-title":"Anim. Behav."},{"key":"2023081107265211100_c15","doi-asserted-by":"publisher","first-page":"491","DOI":"10.1007\/s00359-010-0621-6","article-title":"Adaptive behavior for texture discrimination by the free-flying big brown bat, Eptesicus fuscus","volume":"197","year":"2011","journal-title":"J. Comp. Physiol. A"},{"key":"2023081107265211100_c16","doi-asserted-by":"publisher","first-page":"181","DOI":"10.1207\/s15326969eco0703_2","article-title":"Human echolocation as a basic form of perception and action","volume":"7","year":"1995","journal-title":"Ecol. Psychol."},{"key":"2023081107265211100_c17","unstructured":"C.\n            Ming\n           and \nJ. A.Simmons, \u201c\nTarget geometry estimation using deep neural networks in sonar sensing,\u201d arXiv:2203.15770 (2022)."},{"key":"2023081107265211100_c18","doi-asserted-by":"publisher","first-page":"3433","DOI":"10.1007\/s00034-019-01094-1","article-title":"A survey: Neural network-based deep learning for acoustic event detection","volume":"38","year":"2019","journal-title":"Circuits Syst. Signal Process."},{"key":"2023081107265211100_c19","doi-asserted-by":"publisher","first-page":"206","DOI":"10.1109\/JSTSP.2019.2908700","article-title":"Deep learning for audio signal processing","volume":"13","year":"2019","journal-title":"IEEE J. Sel. Top. Signal Process."},{"key":"2023081107265211100_c20","doi-asserted-by":"publisher","first-page":"420","DOI":"10.1007\/s42979-021-00815-1","article-title":"Deep learning: A comprehensive overview on techniques, taxonomy, applications and research directions","volume":"2","year":"2021","journal-title":"SN Comput. Sci."},{"key":"2023081107265211100_c21","first-page":"443","article-title":"Applications","volume-title":"Deep Learning","year":"2016"},{"key":"2023081107265211100_c22","doi-asserted-by":"publisher","first-page":"3590","DOI":"10.1121\/1.5133944","article-title":"Machine learning in acoustics: Theory and applications","volume":"146","year":"2019","journal-title":"J. Acoust. Soc. Am."},{"key":"2023081107265211100_c23","doi-asserted-by":"publisher","first-page":"3914","DOI":"10.1121\/10.0007284","article-title":"Impact of data augmentation on supervised learning for a moving mid-frequency source","volume":"150","year":"2021","journal-title":"J. Acoust. Soc. Am."},{"key":"2023081107265211100_c24","doi-asserted-by":"publisher","first-page":"279","DOI":"10.1109\/LSP.2017.2657381","article-title":"Deep convolutional neural networks and data augmentation for environmental sound classification","volume":"24","year":"2017","journal-title":"IEEE Signal Process. Lett."},{"key":"2023081107265211100_c25","doi-asserted-by":"publisher","first-page":"101084","DOI":"10.1016\/j.ecoinf.2020.101084","article-title":"Data augmentation approaches for improving animal audio classification","volume":"57","year":"2020","journal-title":"Ecol. Inform."},{"key":"2023081107265211100_c26","doi-asserted-by":"publisher","first-page":"226","DOI":"10.1186\/s12911-022-01942-2","article-title":"On the analysis of data augmentation methods for spectral imaged based heart sound classification using convolutional neural networks","volume":"22","year":"2022","journal-title":"BMC Med. Inf. Decis. Making"},{"key":"2023081107265211100_c27","doi-asserted-by":"publisher","first-page":"107581","DOI":"10.1016\/j.apacoust.2020.107581","article-title":"Spectral images based environmental sound classification using CNN with meaningful data augmentation","volume":"172","year":"2021","journal-title":"Appl. Acoust."},{"key":"2023081107265211100_c28","volume-title":"Echo Signal Processing","year":"2003"},{"key":"2023081107265211100_c29","volume-title":"Principles of Radar and Sonar Signal Processing","year":"2002"},{"key":"2023081107265211100_c30","doi-asserted-by":"publisher","first-page":"547","DOI":"10.1121\/1.1918166","article-title":"Generalized form of the sonar equations","volume":"34","year":"1962","journal-title":"J. Acoust. Soc. Am."},{"key":"2023081107265211100_c31","doi-asserted-by":"crossref","first-page":"257","DOI":"10.1007\/978-1-4614-9146-0","article-title":"Echolocation in air and water","volume-title":"Biosonar","author":"Surlykke","year":"2014"},{"key":"2023081107265211100_c32","first-page":"207","article-title":"Target and background signatures","volume-title":"Principles of Radar and Sonar Signal Processing","year":"2002"},{"key":"2023081107265211100_c33","doi-asserted-by":"crossref","first-page":"319","DOI":"10.1007\/978-1-4615-0312-5_6","article-title":"Spread scattering and propagation","volume-title":"Echo Signal Processing","year":"2003"},{"key":"2023081107265211100_c34","doi-asserted-by":"crossref","first-page":"169","DOI":"10.1007\/978-1-4614-9146-0","article-title":"Localization and classification of targets by echolocating bats and dolphins","volume-title":"Biosonar","author":"Surlykke","year":"2014"},{"key":"2023081107265211100_c35","doi-asserted-by":"publisher","first-page":"623","DOI":"10.2466\/pms.1982.55.2.623","article-title":"Echo perception of shape and texture by sighted subjects","volume":"55","year":"1982","journal-title":"Percept. Mot. Skills"},{"key":"2023081107265211100_c36","doi-asserted-by":"publisher","first-page":"399","DOI":"10.1126\/science.137.3528.399","article-title":"Sonar system of the blind: New research measures their accuracy in detecting the texture, size, and distance of objects \u2018by ear","volume":"137","year":"1962","journal-title":"Science"},{"key":"2023081107265211100_c37","doi-asserted-by":"publisher","first-page":"1333","DOI":"10.1121\/1.398694","article-title":"The acoustic basis for target discrimination by FM echolocating bats","volume":"86","year":"1989","journal-title":"J. Acoust. Soc. Am."},{"key":"2023081107265211100_c38","doi-asserted-by":"crossref","first-page":"177","DOI":"10.1007\/978-1-4612-4356-4_9","article-title":"Biosonar discrimination, recognition, and classification","volume-title":"The Sonar of Dolphins","year":"1993"},{"key":"2023081107265211100_c39","first-page":"35","article-title":"ADC and DAC","volume-title":"The Scientist and Engineer's Guide to Digital Signal Processing","year":"1999"},{"key":"2023081107265211100_c40","doi-asserted-by":"publisher","first-page":"93","DOI":"10.1109\/19.126639","article-title":"Digital time-of-flight measurement for ultrasonic sensors","volume":"41","year":"1992","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"2023081107265211100_c41","first-page":"25","article-title":"Time-frequency transforms","volume-title":"Time-Frequency Transforms for Radar Imaging and Signal Analysis","year":"2002"},{"key":"2023081107265211100_c42","doi-asserted-by":"crossref","first-page":"13","DOI":"10.1007\/3-540-27481-2_2","article-title":"Continuous analysis","volume-title":"Wavelets and Signal Processing: An Application-based Introduction","year":"2005"},{"key":"2023081107265211100_c43","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1007\/978-3-030-87528-2_1","article-title":"Fundamental concepts","volume-title":"Wavelets and Wavelet Transform Systems and Their Applications","year":"2022"},{"key":"2023081107265211100_c44","doi-asserted-by":"publisher","first-page":"2661","DOI":"10.1109\/TSP.2002.804066","article-title":"Generalized Morse wavelets","volume":"50","year":"2002","journal-title":"IEEE Trans. Signal Process."},{"key":"2023081107265211100_c45","first-page":"93","article-title":"Multivariate methods","volume-title":"Introduction to Machine Learning","year":"2014"},{"key":"2023081107265211100_c46","first-page":"228","article-title":"Regularization for deep learning","volume-title":"Deep Learning","year":"2016"},{"key":"2023081107265211100_c47","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1186\/s40537-019-0197-0","article-title":"A survey on image data augmentation for deep learning","volume":"6","year":"2019","journal-title":"J. Big Data"},{"key":"2023081107265211100_c48","first-page":"215","article-title":"Data augmentation on synthetic images for transfer learning using deep CNNs","year":"2018"},{"key":"2023081107265211100_c49","doi-asserted-by":"publisher","first-page":"1765","DOI":"10.1121\/1.4733537","article-title":"Estimation of the acoustic range of bat echolocation for extended targets","volume":"132","year":"2012","journal-title":"J. Acoust. Soc. Am."},{"key":"2023081107265211100_c50","first-page":"2613","article-title":"SpecAugment: A simple data augmentation method for automatic speech recognition","year":"2019"},{"key":"2023081107265211100_c51","unstructured":"H.\n            Zhang\n          , \nM.Cisse, \nY. N.Dauphin, and \nD.Lopez-Paz, \u201c\nmixup: Beyond empirical risk minimization,\u201d arXiv:1710.09412 (2018)."},{"key":"2023081107265211100_c52","doi-asserted-by":"crossref","first-page":"231","DOI":"10.1007\/978-3-030-74478-6","article-title":"Deep learning for audio and music","volume-title":"Multi-Faceted Deep Learning","author":"Benois-Pineau","year":"2021"},{"key":"2023081107265211100_c53","first-page":"330","article-title":"Convolutional networks","volume-title":"Deep Learning","year":"2016"},{"key":"2023081107265211100_c54","doi-asserted-by":"crossref","first-page":"85","DOI":"10.1007\/978-3-319-57550-6_3","article-title":"Convolutional neural networks","volume-title":"Guide to Convolutional Neural Networks","year":"2017"},{"key":"2023081107265211100_c55","unstructured":"S.\n            Ioffe\n           and \nC.Szegedy, \u201c\nBatch normalization: Accelerating deep network training by reducing internal covariate shift,\u201d arXiv:1502.03167 (2015)."},{"key":"2023081107265211100_c56","first-page":"168","article-title":"Deep feedforward networks","volume-title":"Deep Learning","year":"2016"},{"key":"2023081107265211100_c57","first-page":"2472","article-title":"Designing efficient architectures for modeling temporal features with convolutional neural networks","year":"2017"},{"key":"2023081107265211100_c58","doi-asserted-by":"publisher","first-page":"1733","DOI":"10.1109\/TMM.2015.2428998","article-title":"Detection and classification of acoustic scenes and events","volume":"17","year":"2015","journal-title":"IEEE Trans. Multimedia"},{"key":"2023081107265211100_c59","first-page":"179","article-title":"Linear models for classification","volume-title":"Information Science and Statistics, in Pattern Recognition and Machine Learning","year":"2006"},{"key":"2023081107265211100_c60","doi-asserted-by":"publisher","first-page":"2278","DOI":"10.1109\/5.726791","article-title":"Gradient-based learning applied to document recognition","volume":"86","year":"1998","journal-title":"Proc. IEEE"},{"key":"2023081107265211100_c61","first-page":"182","article-title":"Class-conditional data augmentation applied to image classification","volume-title":"Lecture Notes in Computer Science, in Computer Analysis of Images and Patterns","author":"Vento","year":"2019"},{"key":"2023081107265211100_c62","doi-asserted-by":"publisher","first-page":"221","DOI":"10.1109\/TPAMI.2012.59","article-title":"3D convolutional neural networks for human action recognition","volume":"35","year":"2013","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"2023081107265211100_c63","unstructured":"X.\n            Shi\n          , \nZ.Chen, \nH.Wang, \nD.-Y.Yeung, \nW.Wong, and \nW.Woo, \u201c\nConvolutional LSTM network: A machine learning approach for precipitation nowcasting,\u201d arXiv:1506:04214 (2015)."}],"container-title":["The Journal of the Acoustical Society of America"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pubs.aip.org\/asa\/jasa\/article-pdf\/153\/4\/2447\/17019269\/2447_1_10.0017922.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/pubs.aip.org\/asa\/jasa\/article-pdf\/153\/4\/2447\/17019269\/2447_1_10.0017922.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/pubs.aip.org\/asa\/jasa\/article-pdf\/153\/4\/2447\/17019269\/2447_1_10.0017922.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,5,27]],"date-time":"2025-05-27T19:14:41Z","timestamp":1748373281000},"score":1,"resource":{"primary":{"URL":"https:\/\/pubs.aip.org\/jasa\/article\/153\/4\/2447\/2884903\/Convolutional-neural-network-with-data"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,4,1]]},"references-count":63,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2023,4,1]]}},"URL":"https:\/\/doi.org\/10.1121\/10.0017922","relation":{},"ISSN":["0001-4966"],"issn-type":[{"value":"0001-4966","type":"print"}],"subject":[],"published-other":{"date-parts":[[2023,4]]},"published":{"date-parts":[[2023,4,1]]}}}