Dynamic right-context chunked attention masking lets a single zipformer ASR model cover streaming and non-streaming use, nearly closing the accuracy gap with a modest latency increase.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SD 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Unifying Streaming and Non-streaming Zipformer-based ASR
Dynamic right-context chunked attention masking lets a single zipformer ASR model cover streaming and non-streaming use, nearly closing the accuracy gap with a modest latency increase.