A two-stage deep learning framework for lead instrument recognition in polyphonic music featuring Chinese instruments

Jiaxiang Zheng, Moxi Cao, Chongbin Zhang · 2026

Read A two-stage deep learning framework for lead instrument recognition in polyphonic music featuring Chinese instruments by Jiaxiang Zheng, Moxi Cao, Chongbin Zhang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0327442.txt.

Description

Jiaxiang Zheng, Moxi Cao, and Chongbin Zhang built a computer system that can pick out the lead instrument in recordings of Chinese traditional music, reported in PLOS ONE in 2026. In an ensemble, many instruments play at once, and telling them apart by ear or software is hard, especially when two flutes sound nearly alike.

Their method works in two stages: first it separates the mixed audio into single-instrument strands, then a classifier names the instruments present. On a new dataset they built, accuracy and related scores rose clearly over earlier approaches, with the biggest gains on sound-alike pairs such as the dizi and xiao flutes. They offer the tool for preserving and studying Chinese musical heritage.

Catalog record

CREATORJiaxiang Zheng, Moxi Cao, Chongbin Zhang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE86566
LICENSECC_BY
RIGHTSCC BY
SOURCEPLOS

Source: https://doi.org/10.1371/journal.pone.0327442

Browse the full library · About the Tower of Babel