A simple neural network trained on spectrotemporal modulation features performs comparably to pretrained audio embeddings for classifying speech, music, and environmental sounds, without any pretraining.
A critical step in developing powerful audio deep neu- ral networks (DNNs) is converting audio signals into meaning- ful acoustic feature representations
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.SD 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Spectrotemporal Modulation: Efficient and Interpretable Feature Representation for Classifying Speech, Music, and Environmental Sounds
A simple neural network trained on spectrotemporal modulation features performs comparably to pretrained audio embeddings for classifying speech, music, and environmental sounds, without any pretraining.