Abstract
Stem retrieval, the task of matching missing stems to a given audio submix, is a key challenge currently limited by models that discard temporal information. We introduce PHALAR, a contrastive framework achieving a relative accuracy increase of up to 70% over the state-of-the-art while requiring <50% of the parameters and a 7× training speedup. By utilizing a Learned Spectral Pooling layer and a complex-valued head, PHALAR enforces pitch-equivariant and phase-equivariant biases. PHALAR establishes new retrieval state-of-the-art across MoisesDB, Slakh, and ChocoChorales, correlating significantly higher with human coherence judgment than semantic baselines. Finally, zero-shot beat tracking and linear chord probing confirm that PHALAR captures robust musical structures beyond the retrieval task.
Davide Marincione
PhD Student
Michele Mancusi
PostDoctoral Researcher
PhD Student @SapienzaRoma CS | Intern @Musixmatch | Intern @Microsoft | Research Scientist @Sony | Senior Research Scientist @Moises
Giorgio Strano
PhD Student
Luca Cerovaz
Research Intern
Donato Crisostomi
PhD Student
PhD student @ Sapienza, University of Rome | former Applied Science intern @ Amazon Search, Luxembourg | former Research Science intern @ Amazon Alexa, Turin
