Report

Paper explores transfer learning for rhythm game chart generation

SoundPapers shared a paper by Sojeong An and colleagues on music transcription and cross-game transfer for chart generation.

2 Sources, 6h ago, first seen 6h ago

Rhythm games ask the player to act in time with a recording, following a chart: a sequence of actions, each with a time and, depending on the game, a lane, type, or playfield position. Charts are hand-written per recording and difficulty. Automatic generation from audio has been studied with pipeline and end-to-end approaches, and recent work reuses pre-trained audio encoders, but never compares them with one trained from scratch. None of these targets automatic music transcription (AMT), yet most chart times are note onsets, which AMT models locate. We freeze a pre-trained audio encoder, an AMT model in particular, and train only a linear projection and a Transformer decoder on charts. With MuScriptor, this fits human charts more closely than encoders trained from scratch and places more accurate note types than MERT, a music encoder pre-trained without transcription labels. Joint training on three osu! modes raises event F1 on the mode with the fewest charts from 0.492 to 0.506, whil

TLDR

SoundPapers shared “Transfer Learning for Rhythm Game Chart Generation from Music Transcription and Across Games,” a paper by Sojeong An, Nathan Pruyne, Chris Donahue and Dasaem Jeong.

Today's Rank

—

Not ranked yet

Combined views

—

2 Sources, first seen 6h ago

Paper explores transfer learning for rhythm game chart generation

TLDR

Combined views

Combined views

Paper explores transfer learning for rhythm game chart generation

TLDR

Combined views

2 Sources

Combined views

Sentiment

Summary

Sentiment

Summary

Sentiment

Summary

Sentiment

Summary

2 Sources