Recently, the speech community is seeing a significant trend of moving from deep neural
network based hybrid modeling to end-to-end (E2E) modeling for automatic speech …

We introduce the Universal Speech Model (USM), a single large model that performs
automatic speech recognition (ASR) across 100+ languages. This is achieved by pre …

Self-attention (SA) is a critical component of Transformer neural networks that have
succeeded in automatic speech recognition (ASR). In this paper, we analyze the role of SA …

In this paper, we present an in-depth study on online attention mechanisms and distillation
techniques for dual-mode (ie, joint online and offline) ASR using the Conformer Transducer …

This work studies the use of attention masking in transformer transducer based speech
recognition for building a single configurable model for different deployment scenarios. We …

Streaming speech recognition architectures are employed for low-latency, real-time
applications. Such architectures are often characterized by their causality. Causal …

The term “automatic speech recognition” refers to the procedure by which an auditory signal
of spoken words can be converted into text. Voice recognition is another term that may be …

The state-of-the-art conformer used in automatic speech recognition combines feed-forward,
convolution and multi-headed self-attention layers in a single model that is trained end-to …

In this paper, we propose the first successful implementation of associated learning (AL) to
automatic speech recognition (ASR). AL has been shown to provide better label noise …

A method for an automated speech recognition (ASR) model for unifying streaming and non-
streaming speech recognition including receiving a sequence of acoustic frames. The …