As the computational requirements for machine learning systems and the s...
We develop an algorithm which can learn from partially labeled and
unseg...
Finite-state transducers (FSTs) are frequently used in speech recognitio...
Secure multi-party computation (MPC) allows parties to perform computati...
The decade from 2010 to 2020 saw remarkable improvements in automatic sp...
Machine intelligence can develop either directly from experience or by
i...
Machine learning systems typically assume that the distributions of trai...
Machine-learning systems such as self-driving cars or virtual assistants...
Machine-learning models contain information about the data they were tra...
One of the most effective approaches to improving the performance of a
m...
We introduce a framework for automatic differentiation with weighted
fin...
Machine learning models leak information about their training data every...
We study training a single acoustic model for multiple languages with th...
Pseudo-labeling has recently shown promise in end-to-end automatic speec...
For sequence transduction tasks like speech recognition, a strong struct...
We design an online end-to-end speech recognition system based on Time-D...
Secure multiparty computations enable the distribution of so-called shar...
Good data stewardship requires removal of data at the request of the dat...
The transcriptions used to train an Automatic Speech Recognition (ASR) s...
Contextual bandits are online learners that, given an input, select an a...
We revisit self-training in the context of end-to-end speech recognition...
We propose a direct-to-word sequence model with a dynamic lexicon. Our w...
We propose a fully convolutional sequence-to-sequence encoder architectu...
We introduce a new beam search decoder that is fully differentiable, mak...
This paper introduces wav2letter++, the fastest open-source deep learnin...
We propose a single neural network architecture for two tasks: on-line
k...
Deep learning has dramatically improved the performance of speech recogn...
We show that an end-to-end deep learning approach can be used to recogni...
We present a state-of-the-art speech recognition system developed using
...