A per-layer gate predictor using speaker and acoustic-event embeddings dynamically prunes an OWSM speech model, cutting encoder GFLOPs by 56.7 while improving Europarl-ST BLEU by about 26% relative.
Colld: Contrastive layer-to-layer distillation for compressing multilingual pre-trained speech encoders,
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
eess.AS 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Context-Driven Dynamic Pruning for Large Speech Foundation Models
A per-layer gate predictor using speaker and acoustic-event embeddings dynamically prunes an OWSM speech model, cutting encoder GFLOPs by 56.7 while improving Europarl-ST BLEU by about 26% relative.