New AI Model RPPNet Generates Melodies Using Music Psychology Principles

Researchers have developed RPPNet, a novel two-stage deep learning architecture for generating melodies with improved long-term structure. Unlike existing models that rely on fixed bar lines, RPPNet generates variable-length Rhythm-Pitch Primitive (RPP) sequences, which are then decoded into musical notes. The grouping of these RPPs is informed by principles of music psychology, including acoustic cues and perceptual similarity. Experiments indicate that RPPNet produces melodies with superior long-term structure and musicality compared to current methods. AI

IMPACT This research could lead to more musically coherent and structurally sound AI-generated music by incorporating principles of human perception.

RANK_REASON The item is an academic paper detailing a new model and its experimental results. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• arXiv
• CatalyzeX Code Finder for Papers
• Connected Papers
• CORE Recommender
• DagsHub
• Gotit.pub
• Hugging Face
• Influence Flower
• Litmaps
• Rhythm-Pitch Primitive
• RPPNet
• ScienceCast
• scite Smart Citations

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →