In this paper, we explore reducing computational latency of the 2-pass cascaded encoder model [1]. Specifically, we experiment with reducing the size of the ...

Reducing computational latency of the 2-pass cascaded encoder model by reducing the size of the causal 1st-pass and adding capacity to the non-causal ...

ABSTRACT. In this paper, we explore reducing computational latency of the 2- pass cascaded encoder model [1]. Specifically, we experiment with.

The shared cascaded encoder consists of 12 conformer layers; the first 5 have access to three frames of right context each, for a total of 3 × 5 × 60ms = 900ms ...

Dec 22, 2023 · In this paper, we explore various frame rate reduction schemes on the two-pass cascaded encoder model to improve its ef- ficiency without ...

A streaming on-device end-to-end model surpassing server-side conventional model quality and latency. ... Improving the latency and quality of cascaded encoders.

Improving the latency and quality of cascaded encoders. TN Sainath, Y He, A ... Improving confidence estimation on out-of-domain data for end-to-end ...

IMPROVING THE LATENCY AND QUALITY OF CASCADED ENCODERS · SPE-73.1: Transducer-Based Streaming Deliberation For Cascaded Encoders · SPE-73.2: IMPROVING THE LATENCY ...

Jan 17, 2024 · Gruenstein, Anmol Gulati, et al., “Improving the latency and quality of cascaded encoders,” in ICASSP 2022-2022 IEEE International Con-.

In this paper, we propose a novel decoding al- gorithm for streaming End-to-end (E2E) auto- matic speech recognition (ASR) models, the double decoder.