Reproduce “Exploring Definitions of Quality and Diversity in Sonic Measurement Spaces” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Björn Þór Jónsson, Çağrı Erdem, Stefano Fasciani, Kyrre Glette
The authors use unsupervised sound representations to organize an evolutionary search for synthesized sounds, periodically updating a two-dimensional PCA map as new sounds are found. In five runs, Dynamic PCA reached a diversity score of 0.514 ± 0.040, versus 0.148 ± 0.004 for Manual BD, measured as average pairwise cosine distance between archived feature vectors. Manual BD had higher coverage and grid mean fitness; this computational result does not establish perceived diversity.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.