1 paper
Yannis Vasilakis, Rachel Bittner, Johan Pauwels
Joint audio-text models are widely used for music retrieval, yet they struggle with semantic phenomena such as negation. Negation is fundamental for distinguishing the absence (or…