BLR-MoE, which adds language-specific attention experts, expert pruning, and router fine-tuning to the LR-MoE architecture, reduces WER by 16.09% relative on a 10,000-hour multilingual ASR benchmark.
Joint ctc-attention based end-to-end speech recognition using multi-task learning,
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.CL 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
BLR-MoE: Boosted Language-Routing Mixture of Experts for Domain-Robust Multilingual E2E ASR
BLR-MoE, which adds language-specific attention experts, expert pruning, and router fine-tuning to the LR-MoE architecture, reduces WER by 16.09% relative on a 10,000-hour multilingual ASR benchmark.