Bidirectional recurrent attention with Direction Dropout matches or exceeds multi-head attention accuracy in a Conformer-Transducer ASR system while increasing throughput by up to 44 percent.
However, Transformers, especially their multi-head attention (MHA) component, are ill-suited for long- form ASR due to quadratic time/memory complexity in se- quence length
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
-
Accurate, fast, cheap: Choose three. Replacing Multi-Head-Attention with Bidirectional Recurrent Attention for Long-Form ASR
Bidirectional recurrent attention with Direction Dropout matches or exceeds multi-head attention accuracy in a Conformer-Transducer ASR system while increasing throughput by up to 44 percent.