2 papers
cs.SD2025
Assessing the Alignment of Audio Representations with Timbre Similarity Ratings
Haokun Tian, Stefan Lattner, Charalampos Saitis
Psychoacoustical so-called "timbre spaces" map perceptual similarity ratings of instrument sounds onto low-dimensional embeddings via multidimensional scaling, but suffer from scal…
cs.SD2025
Hybrid Losses for Hierarchical Embedding Learning
Haokun Tian, Stefan Lattner, Brian McFee +1
In traditional supervised learning, the cross-entropy loss treats all incorrect predictions equally, ignoring the relevance or proximity of wrong labels to the correct answer. By l…