{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,14]],"date-time":"2026-06-14T08:12:33Z","timestamp":1781424753454,"version":"3.54.1"},"reference-count":46,"publisher":"Springer Science and Business Media LLC","issue":"5","license":[{"start":{"date-parts":[[2022,6,18]],"date-time":"2022-06-18T00:00:00Z","timestamp":1655510400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2022,6,18]],"date-time":"2022-06-18T00:00:00Z","timestamp":1655510400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["SN COMPUT. SCI."],"abstract":"<jats:title>Abstract<\/jats:title><jats:p>Research on automatic music generation lacks consideration of the originality of musical outputs, creating risks of plagiarism and\/or copyright infringement. We present the <jats:italic>originality report<\/jats:italic>\u2014a set of analyses that is parameterised by a \u201csimilarity score\u201d\u2014for measuring the extent to which an algorithm copies from the input music. First, we construct a baseline, to determine the extent to which human composers borrow from themselves and each other in some existing music corpus. Second, we apply a similar analysis to musical outputs of runs of MAIA Markov and Music Transformer generation algorithms, and compare the results to the baseline. Third, we investigate how originality varies as a function of Transformer\u2019s training epoch. Fourth, we demonstrate the originality report with a different \u201csimilarity score\u201d based on symbolic fingerprinting, encompassing music with more complex, expressive timing information. Results indicate that the originality of Transfomer\u2019s output is below the 95% confidence interval of the baseline. Musicological interpretation of the analyses shows that the Transformer model obtained via the conventional stopping criteria produces single-note repetition patterns, resulting in outputs of low quality and originality, while in later training epochs, the model tends to overfit, producing copies of excerpts of input pieces. Even with a larger data set, the same copying issues still exist. Thus, we recommend the originality report as a new means of evaluating algorithm training processes and outputs in future, and question the reported success of language-based deep learning models for music generation. Supporting materials (data sets and code) are available via <jats:ext-link xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" ext-link-type=\"uri\" xlink:href=\"https:\/\/osf.io\/96emr\/\">https:\/\/osf.io\/96emr\/<\/jats:ext-link>.<\/jats:p>","DOI":"10.1007\/s42979-022-01220-y","type":"journal-article","created":{"date-parts":[[2022,6,18]],"date-time":"2022-06-18T15:02:32Z","timestamp":1655564552000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":9,"title":["Measuring When a Music Generation Algorithm Copies Too Much: The Originality Report, Cardinality Score, and Symbolic Fingerprinting by Geometric Hashing"],"prefix":"10.1007","volume":"3","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8709-8829","authenticated-orcid":false,"given":"Zongyu","family":"Yin","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Federico","family":"Reuben","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3146-5401","authenticated-orcid":false,"given":"Susan","family":"Stepney","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7880-5093","authenticated-orcid":false,"given":"Tom","family":"Collins","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2022,6,18]]},"reference":[{"key":"1220_CR1","unstructured":"Arzt A, B\u00f6ck S, Widmer G. Fast identification of piece and score position via symbolic fingerprinting. In: ISMIR; 2012. pp. 433\u20138."},{"key":"1220_CR2","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1016\/j.compag.2018.08.013","volume":"153","author":"JGA Barbedo","year":"2018","unstructured":"Barbedo JGA. Impact of dataset size and variety on the effectiveness of deep learning and transfer learning for plant disease classification. Comput Electron Agric. 2018;153:46\u201353.","journal-title":"Comput Electron Agric"},{"key":"1220_CR3","doi-asserted-by":"crossref","unstructured":"Bengio Y. Practical recommendations for gradient-based training of deep architectures. In: Neural networks: tricks of the trade. Berlin: Springer; 2012. pp. 437\u201378.","DOI":"10.1007\/978-3-642-35289-8_26"},{"key":"1220_CR4","unstructured":"Carlini N, Tramer F, Wallace E, Jagielski M, Herbert-Voss A, Lee K, Roberts A, Brown T, Song D, Erlingsson U, et\u00a0al. Extracting training data from large language models. 2020. arXiv:2012.07805."},{"key":"1220_CR5","unstructured":"Collins T. Improved methods for pattern discovery in music, with applications in automated stylistic composition. Ph.D. thesis, The Open University. 2011."},{"key":"1220_CR6","doi-asserted-by":"crossref","unstructured":"Collins T, Arzt A, Frostel H, Widmer G. Using geometric symbolic fingerprinting to discover distinctive patterns in polyphonic music corpora. In: Computational music analysis, pp. 445\u201374. Springer. 2016.","DOI":"10.1007\/978-3-319-25931-4_17"},{"key":"1220_CR7","unstructured":"Collins T, B\u00f6ck S, Krebs F, Widmer G. Bridging the audio-symbolic gap: the discovery of repeated note content directly from polyphonic music audio. In: 53rd international conference: semantic audio. Audio Engineering Society. 2014."},{"key":"1220_CR8","unstructured":"Collins T, Coulon C. MAIA Util: an NPM package for bridging web audio with music-theoretic concepts. In: Proceedings of the web audio conference, pp. 47\u201352. Trondheim, Norway. 2019."},{"key":"1220_CR9","doi-asserted-by":"publisher","first-page":"2","DOI":"10.5920\/JCMS.2017.02","volume":"1","author":"T Collins","year":"2017","unstructured":"Collins T, Laney R. Computer-generated stylistic compositions with long-term repetitive and phrasal structure. J Creative Music Syst. 2017;1:2. https:\/\/doi.org\/10.5920\/JCMS.2017.02.","journal-title":"J Creative Music Syst"},{"issue":"1","key":"1220_CR10","first-page":"16","volume":"30","author":"T Collins","year":"2016","unstructured":"Collins T, Laney R, Willis A, Garthwaite PH. Developing and evaluating computational models of musical style. AI EDAM. 2016;30(1):16\u201343.","journal-title":"AI EDAM"},{"issue":"5","key":"1220_CR11","doi-asserted-by":"publisher","first-page":"547","DOI":"10.3233\/IDA-2010-0438","volume":"14","author":"D Conklin","year":"2010","unstructured":"Conklin D. Discovery of distinctive patterns in music. Intell Data Anal. 2010;14(5):547\u201354.","journal-title":"Intell Data Anal"},{"issue":"1","key":"1220_CR12","doi-asserted-by":"publisher","first-page":"51","DOI":"10.1080\/09298219508570672","volume":"24","author":"D Conklin","year":"1995","unstructured":"Conklin D, Witten IH. Multiple viewpoint systems for music prediction. J New Music Res. 1995;24(1):51\u201373.","journal-title":"J New Music Res"},{"key":"1220_CR13","volume-title":"Computer models of musical creativity","author":"D Cope","year":"2005","unstructured":"Cope D. Computer models of musical creativity. Cambridge: MIT Press; 2005."},{"key":"1220_CR14","unstructured":"Donahue C, Mao HH, Li YE, Cottrell GW, McAuley J. LakhNES: improving multi-instrumental music generation with cross-domain pre-training. In: ISMIR; 2019. pp. 685\u201392."},{"key":"1220_CR15","doi-asserted-by":"crossref","unstructured":"Dong HW, Hsiao WY, Yang LC, Yang YH. Musegan: multi-track sequential generative adversarial networks for symbolic music generation and accompaniment. In: Thirty-Second AAAI conference on artificial intelligence. 2018.","DOI":"10.1609\/aaai.v32i1.11312"},{"key":"1220_CR16","unstructured":"Elgammal A, Liu B, Elhoseiny M, Mazzone M. Can: creative adversarial networks generating \u201cart\u201d by learning about styles and deviating from style norms. In: 8th international conference on computational creativity, ICCC 2017. Georgia Institute of Technology. 2017."},{"key":"1220_CR17","doi-asserted-by":"crossref","unstructured":"Graves A, Mohamed Ar, Hinton G, Speech recognition with deep recurrent neural networks. In: 2013 IEEE international conference on acoustics, speech and signal processing, IEEE; 2013. pp. 6645\u201349.","DOI":"10.1109\/ICASSP.2013.6638947"},{"key":"1220_CR18","unstructured":"Hawthorne C, Stasyuk A, Roberts A, Simon I, Huang CZA, Dieleman S, Elsen E, Engel J, Eck D. Enabling factorized piano music modeling and generation with the MAESTRO dataset. In: international conference on learning representations. 2019."},{"key":"1220_CR19","unstructured":"Huang CZA, Vaswani A, Uszkoreit J, Simon I, Hawthorne C, Shazeer N, Dai AM, Hoffman MD, Dinculescu M, Eck D. Music transformer: generating music with long-term structure. In: ICLR. 2018."},{"key":"1220_CR20","unstructured":"Janssen B, Collins T, Ren IY. Algorithmic ability to predict the musical future: Datasets and evaluation. In: ISMIR; 2019. pp. 208\u201315."},{"key":"1220_CR21","unstructured":"Kingma DP, Ba J, Adam: A method for stochastic optimization. 2014. arXiv:1412.6980."},{"key":"1220_CR22","doi-asserted-by":"publisher","DOI":"10.1093\/acprof:oso\/9780195317138.001.0001","volume-title":"Generalized musical intervals and transformations","author":"D Lewin","year":"2007","unstructured":"Lewin D. Generalized musical intervals and transformations. Oxford: Oxford University Press; 2007. (originally published by Yale University Press, New Haven, 1987)."},{"issue":"2","key":"1220_CR23","doi-asserted-by":"publisher","first-page":"91","DOI":"10.1023\/B:VISI.0000029664.99615.94","volume":"60","author":"DG Lowe","year":"2004","unstructured":"Lowe DG. Distinctive image features from scale-invariant keypoints. Int J Comput Vis. 2004;60(2):91\u2013110.","journal-title":"Int J Comput Vis"},{"issue":"2","key":"1220_CR24","doi-asserted-by":"publisher","first-page":"121","DOI":"10.1080\/09298210600834961","volume":"35","author":"D Meredith","year":"2006","unstructured":"Meredith D. The ps13 pitch spelling algorithm. J New Music Res. 2006;35(2):121\u201359.","journal-title":"J New Music Res"},{"issue":"4","key":"1220_CR25","doi-asserted-by":"publisher","first-page":"321","DOI":"10.1076\/jnmr.31.4.321.14162","volume":"31","author":"D Meredith","year":"2002","unstructured":"Meredith D, Lemstr\u00f6m K, Wiggins GA. Algorithms for discovering repeated patterns in multidimensional representations of polyphonic music. J New Music Res. 2002;31(4):321\u201345.","journal-title":"J New Music Res"},{"issue":"1","key":"1220_CR26","doi-asserted-by":"publisher","first-page":"257","DOI":"10.1177\/102986490901300111","volume":"13","author":"D M\u00fcllensiefen","year":"2009","unstructured":"M\u00fcllensiefen D, Pendzich M. Court decisions on music plagiarism and the predictive value of similarity algorithms. Musicae Sci. 2009;13(1):257\u201395.","journal-title":"Musicae Sci"},{"key":"1220_CR27","unstructured":"M\u00fcller R, Kornblith S, Hinton GE. When does label smoothing help? In: Advances in neural information processing systems; 2019. pp. 4694\u2013703."},{"key":"1220_CR28","unstructured":"Neely A. Why the Katy Perry\/Flame lawsuit makes no sense. 2020. https:\/\/www.youtube.com\/watch?v=0ytoUuO-qvg. Accessed 30 Oct 2020."},{"key":"1220_CR29","first-page":"1","volume":"2018","author":"S Oore","year":"2018","unstructured":"Oore S, Simon I, Dieleman S, Eck D, Simonyan K. This time with feeling: learning expressive musical performance. Neural Comput Appl. 2018;2018:1\u201313.","journal-title":"Neural Comput Appl"},{"key":"1220_CR30","unstructured":"Otzen E. Six seconds that shaped 1500 songs. 2020. https:\/\/www.bbc.co.uk\/news\/magazine-32087287. Accessed 30 Oct 2020."},{"key":"1220_CR31","unstructured":"Pachet F, Papadopoulos A, Roy P. Sampling variations of sequences for structured music generation. In: ISMIR; 2017. pp. 167\u201373."},{"key":"1220_CR32","unstructured":"Radford A, Metz L, Chintala S. Unsupervised representation learning with deep convolutional generative adversarial networks. 2015. arXiv:1511.06434."},{"key":"1220_CR33","unstructured":"Roberts A, Engel J, Raffel C, Hawthorne C, Eck D. A hierarchical latent vector model for learning long-term structure in music. In: International conference on machine learning. PMLR; 2018. pp. 4364\u201373."},{"key":"1220_CR34","unstructured":"Ruiz A, Simon I. My only problem with Magenta\u2019s Transformer. Magenta Discuss Google Group. 2020. https:\/\/groups.google.com\/a\/tensorflow.org\/g\/magenta-discuss\/c\/Oxiq-Gdaavk\/m\/uHIsQZKtBwAJ. Accessed 30 Oct 2020."},{"key":"1220_CR35","first-page":"93","volume":"40","author":"E Selfridge-Field","year":"1999","unstructured":"Selfridge-Field E. Conceptual and representational issues in melodic comparison. Comput Musicol. 1999;40:93.","journal-title":"Comput Musicol"},{"key":"1220_CR36","first-page":"1","volume":"25","author":"I Stav","year":"2014","unstructured":"Stav I. Musical plagiarism: a true challenge for the copyright law. DePaul J Art Tech Intell Prop L. 2014;25:1.","journal-title":"DePaul J Art Tech Intell Prop L"},{"key":"1220_CR37","doi-asserted-by":"crossref","unstructured":"Sturm BL, Iglesias M, Ben-Tal O, Miron M, G\u00f3mez E. Artificial intelligence and music: open questions of copyright law and engineering praxis. In: Arts, vol.\u00a08, p.\u00a0115. Multidisciplinary Digital Publishing Institute. 2019.","DOI":"10.3390\/arts8030115"},{"issue":"4","key":"1220_CR38","doi-asserted-by":"publisher","first-page":"27","DOI":"10.2307\/3679551","volume":"13","author":"PM Todd","year":"1989","unstructured":"Todd PM. A connectionist approach to algorithmic composition. Comput Music J. 1989;13(4):27\u201343.","journal-title":"Comput Music J"},{"key":"1220_CR39","unstructured":"Ukkonen E, Lemstr\u00f6m K, M\u00e4kinen V, Geometric algorithms for transposition invariant content-based music retrieval. In: Proceedings of the international symposium on music information retrieval; 2003. pp. 193\u20139."},{"key":"1220_CR40","unstructured":"Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser \u0141, Polosukhin I. Attention is all you need. In: Advances in neural information processing systems; 2017. pp. 5998\u20136008."},{"issue":"3","key":"1220_CR41","doi-asserted-by":"publisher","first-page":"207","DOI":"10.1080\/09298215.2016.1232412","volume":"45","author":"A Volk","year":"2016","unstructured":"Volk A, Chew E, Margulis EH, Anagnostopoulou C. Music similarity: concepts, cognition and computation. J New Music Res. 2016;45(3):207\u20139.","journal-title":"J New Music Res"},{"key":"1220_CR42","unstructured":"Wang ALC, Smith\u00a0III JO. System and methods for recognizing sound and music signals in high noise and distortion (2012), patent US 8,190,435 B2. Continuation of provisional application from 2000. 2012."},{"issue":"7","key":"1220_CR43","doi-asserted-by":"publisher","first-page":"1336","DOI":"10.1109\/TIFS.2015.2407698","volume":"10","author":"X Wang","year":"2015","unstructured":"Wang X, Pang K, Zhou X, Zhou Y, Li L, Xue J. A visual model-based perceptual image hash for content authentication. IEEE Trans Inf Forensics Secur. 2015;10(7):1336\u201349.","journal-title":"IEEE Trans Inf Forensics Secur"},{"issue":"2","key":"1220_CR44","doi-asserted-by":"publisher","first-page":"129","DOI":"10.1016\/S0004-3702(03)00016-X","volume":"146","author":"G Widmer","year":"2003","unstructured":"Widmer G. Discovering simple rules in complex data: A meta-learning algorithm and some surprising musical discoveries. Artif Intell. 2003;146(2):129\u201348.","journal-title":"Artif Intell"},{"key":"1220_CR45","volume-title":"Twentieth century music: its evolution from the end of the harmonic era into the present era of sound","author":"P Yates","year":"1968","unstructured":"Yates P. Twentieth century music: its evolution from the end of the harmonic era into the present era of sound. New York: Allen & Unwin; 1968."},{"key":"1220_CR46","unstructured":"Zauner C. Implementation and benchmarking of perceptual image hash functions. In: Master\u2019s thesis, Upper Austria University of Applied Sciences. 2010."}],"container-title":["SN Computer Science"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s42979-022-01220-y.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s42979-022-01220-y\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s42979-022-01220-y.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,9,19]],"date-time":"2022-09-19T19:08:24Z","timestamp":1663614504000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s42979-022-01220-y"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,6,18]]},"references-count":46,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2022,9]]}},"alternative-id":["1220"],"URL":"https:\/\/doi.org\/10.1007\/s42979-022-01220-y","relation":{},"ISSN":["2661-8907"],"issn-type":[{"value":"2661-8907","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,6,18]]},"assertion":[{"value":"11 September 2021","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"19 May 2022","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"18 June 2022","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The last author is a co-founder of the music collective Music Artificial Intelligence Algorithms, Inc. As such, he has an interest in the strong performance of the MAIA Markov algorithm. However, such interest does not affect using the introduced method to produce objective results. Apart from this, the authors have no relevant financial or non-financial interests to disclose.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of Interest"}},{"value":"Not applicable.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethics Approval"}},{"value":"All authors consent to the publication of this work.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for Publication"}},{"value":"Not applicable","order":5,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent to Participate"}}],"article-number":"340"}}