Self-supervised learning of speech representations has been a very active research area
but most work is focused on a single domain such as read audio books for which there exist …

Speech language models (SpeechLMs) process and generate acoustic data only, without
textual supervision. In this work, we propose TWIST, a method for training SpeechLMs using …

Audio is an essential part of our life, but creating it often requires expertise and is time-
consuming. Research communities have made great progress over the past year advancing …

In this paper, we propose TitaNet, a novel neural network architecture for extracting speaker
representations. We employ 1D depth-wise separable convolutions with Squeeze-and …

Is pushing numbers on a single benchmark valuable in automatic speech recognition?
Research results in acoustic modeling are typically evaluated based on performance on a …

Abstract Language-users reduce words in predictable contexts. Previous research indicates
that reduction may be stored in lexical representation if a word is often reduced. Because …

It is natural and efficient to use human natural language (NL) directly to instruct robot task
executions without prior user knowledge of instruction patterns. Currently, NL-instructed …

Recent results in end-to-end automatic speech recognition have demonstrated the efficacy
of pseudo-labeling for semi-supervised models trained both with Connectionist Temporal …

This paper is a study of performance-efficiency trade-offs in pre-trained models for automatic
speech recognition (ASR). We focus on wav2vec 2.0, and formalize several architecture …

Statistical learning (SL) is considered a cornerstone of cognition. While decades of research
have unveiled the remarkable breadth of structures that participants can learn from statistical …