{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/an-end-to-end-neural-network-for-polyphonic","title":"An End-to-End Neural Network for Polyphonic Piano Music Transcription","arxiv_id":"1508.01774","date":"2015-08-07","proceeding":null,"authors":["Siddharth Sigtia","Emmanouil Benetos","Simon Dixon"],"abstract":"We present a supervised neural network model for polyphonic piano music\ntranscription. The architecture of the proposed model is analogous to speech\nrecognition systems and comprises an acoustic model and a music language model.\nThe acoustic model is a neural network used for estimating the probabilities of\npitches in a frame of audio. The language model is a recurrent neural network\nthat models the correlations between pitch combinations over time. The proposed\nmodel is general and can be used to transcribe polyphonic music without\nimposing any constraints on the polyphony. The acoustic and language model\npredictions are combined using a probabilistic graphical model. Inference over\nthe output variables is performed using the beam search algorithm. We perform\ntwo sets of experiments. We investigate various neural network architectures\nfor the acoustic models and also investigate the effect of combining acoustic\nand music language model predictions using the proposed architecture. We\ncompare performance of the neural network based acoustic models with two\npopular unsupervised acoustic models. Results show that convolutional neural\nnetwork acoustic models yields the best performance across all evaluation\nmetrics. We also observe improved performance with the application of the music\nlanguage models. Finally, we present an efficient variant of beam search that\nimproves performance and reduces run-times by an order of magnitude, making the\nmodel suitable for real-time applications.","url_abs":"http://arxiv.org/abs/1508.01774v2","url_pdf":"http://arxiv.org/pdf/1508.01774v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"an-end-to-end-neural-network-for-polyphonic","repo_url":"https://github.com/9552nZ/SmartSheetMusic","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"an-end-to-end-neural-network-for-polyphonic","repo_url":"https://github.com/jsleep/wav2mid","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"music-transcription","task_name":"Music Transcription"},{"task_slug":"speech-recognition","task_name":"Speech Recognition"},{"task_slug":"speech-recognition-1","task_name":"speech-recognition"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}