New method uses AI embeddings to evaluate piano performance

Researchers have developed a new method for evaluating expressive MIDI piano performances by integrating contextual embeddings from self-supervised symbolic music models like Aria and CLaMP3. This approach moves beyond traditional attribute statistics by considering inter-note dependencies and offers a more robust way to assess performance similarity. The study found that these models can serve as perceptual proxies, aligning with human ratings, and adapted Kernel Audio Distance for symbolic music to measure distributional similarity without requiring note alignment. AI

IMPACT This research could lead to more sophisticated AI-driven tools for music generation and performance analysis.

RANK_REASON The item is an academic paper detailing a new methodology for evaluating MIDI piano performances using AI models and metrics. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• Aria
• arXiv
• CLaMP3
• DagsHub
• Hugging Face
• Kernel Audio Distance
• Pearson product-moment correlation coefficient
• Pereval

• paper
• other

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