Master research Internship Internship report Segmentation and recognition of symbols for printed and handwritten music scores
نویسنده
چکیده
An OMR system objective is to transcribe an image of a music score into a machine readable format. A music grammar can be used in order to model the structural knowledge of music scores. Music scores can have a high density of symbols and the different preprocessing step can lead to wrongly segmented components. However, we don’t want to tell the grammar how a music symbol should be segmented. Statistical model like convolutional neural network are much more suited for localizing and classifying music symbols In this report, we propose a way to combine a music grammar with a convolutional neural network architecture capable of doing localization and classification of music symbols. The grammar is used to generate small contextual zone that will be then fed to the neural network. This grammar can then also be used to bootstrap a dataset generation for the training of neural networks. We demonstrate the use of our model on the concrete task of recognizing accidentals in a music score and obtain 93.4% of localization accuracy and 96.2% of classification accuracy.
منابع مشابه
Staff and graphical primitive segmentation in old handwritten music scores
In document analysis field, Optical Musical Recognition is a mature area in printed scores, whereas few research works have been done in handwritten ones. The difficulties in handwritten scores are increased if we work in old documents, because of paper degradation and the lack of a standard in musical notation. In this paper we propose a method to segment staff and graphical primitives in old ...
متن کاملStudy and Analysis for Development of an Efficient OCR for Printed and Handwritten ODIA Documents: A Survey
The OCR (optical character recognition) is the process of translating the hand written or printed text into a format that is understood by the machine for the purpose of editing, searching and indexing. Preprocessing, segmentation, features extraction, classification and post processing are the main phases of any OCR system and these specific fields are in use today. For all these tasks the seg...
متن کاملOff-line Arabic Handwritten Recognition Using a Novel Hybrid HMM-DNN Model
In order to facilitate the entry of data into the computer and its digitalization, automatic recognition of printed texts and manuscripts is one of the considerable aid to many applications. Research on automatic document recognition started decades ago with the recognition of isolated digits and letters, and today, due to advancements in machine learning methods, efforts are being made to iden...
متن کامل