Gauri Bhagwat · Baselines · Published August 30, 2023
Text Line Detection in Printed Music Books
Layout Analysis
Description
This model is trained on images from 16th-century printed Music Books from the KBR collection in Brussels, Belgium. The layout of the pages used for this training is landscape. It specializes in identifying text regions within the images, such as lyrics, page numbers, and other textual content. It doesn't focus on musical notations or decorative elements like staff notes or fancy letters.
Its purpose is to help transcribing lyrics from those pages easier.
The image displayed is for the purpose of representation and is not taken from the training dataset.
Open in Transkribus
Very low loss2.73% loss
Loss indicates how far the predicted text regions deviate from the ground truth (lower is better). This model achieved 2.73% on its validation set. A loss below 10% generally indicates reliable baseline detection.
Layout detection quality depends heavily on your document's structure. Pages with columns, marginalia, or non-standard layouts may produce different results.
Words10,179
Lines1,693
Training Pages217
Model ID54770