A combination of reversible residual blocks and 8-bit optimizer state quantization trains deep speaker embedding extractors with up to 16.2x less GPU memory and comparable accuracy.
X-vectors: Robust dnn embeddings for speaker recognition,
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
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Memory-Efficient Training for Deep Speaker Embedding Learning in Speaker Verification
A combination of reversible residual blocks and 8-bit optimizer state quantization trains deep speaker embedding extractors with up to 16.2x less GPU memory and comparable accuracy.