A 38MB respiratory-sound classifier using grouped spectrograms, deep clustering, and group-mix contrastive learning reports an ICBHI2017 score of 63.26% (Sp 82.06%, Se 44.47%) without pretrained weights.
Who global progress report on tuberculosis elimination,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SD 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
CycleGuardian: A Framework for Automatic RespiratorySound classification Based on Improved Deep clustering and Contrastive Learning
A 38MB respiratory-sound classifier using grouped spectrograms, deep clustering, and group-mix contrastive learning reports an ICBHI2017 score of 63.26% (Sp 82.06%, Se 44.47%) without pretrained weights.