A small trained 'seed' speech encoder can be unfolded to several logical depths by repeating shared layers, matching independently trained models with up to 35% parameter reduction.
Experimental Setup Models and Data.For supervised learning, we take Conformer AED system configured with the ESPnet [45] recipe2
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
-
Unfolding A Few Structures for The Many: Memory-Efficient Compression of Conformer and Speech Foundation Models
A small trained 'seed' speech encoder can be unfolded to several logical depths by repeating shared layers, matching independently trained models with up to 35% parameter reduction.