English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The first step to apply deep learning techniques for symbolic musicunderstanding is to transform musical pieces (mainly in MIDI format) intosequences of predefined tokens like note pitch, note velocity, and chords.Subsequently, the sequences are fed into a neural sequence model to accomplishspecific tasks. Music sequences exhibit strong correlations between adjacentelements, making them prime candidates for N-gram techniques from NaturalLanguage Processing (NLP). Consider classical piano music: specific melodiesmight recur throughout a piece, with subtle variations each time. In thispaper, we propose a novel method, NG-Midiformer, for understanding symbolicmusic sequences that leverages the N-gram approach. Our method involves firstprocessing music pieces into word-like sequences with our proposed unsupervisedcompoundation, followed by using our N-gram Transformer encoder, which caneffectively incorporate N-gram information to enhance the primary encoder partfor better understanding of music sequences. The pre-training process onlarge-scale music datasets enables the model to thoroughly learn the N-graminformation contained within music sequences, and subsequently apply thisinformation for making inferences during the fine-tuning stage. Experiment onvarious datasets demonstrate the effectiveness of our method and achievedstate-of-the-art performance on a series of music understanding downstreamtasks. The code and model weights will be released athttps://github.com/CinqueOrigin/NG-Midiformer.

N-Gram Unsupervised Compoundation and Feature Injection for Better Symbolic Music Understanding