Deep learning is an evolving expanse of machine learning. Machine learning is observing
its neoteric span as deep learning is steadily becoming the pioneer in this field. With the …

Sequence-to-sequence models have shown success in end-to-end speech recognition.
However these models have only used shallow acoustic encoder networks. In our work, we …

A key problem in structured output prediction is enabling direct optimization of the task
reward function that matters for test evaluation. This paper presents a simple and …

This paper presents the Imputer, a neural sequence model that generates output sequences
iteratively via imputations. The Imputer is an iterative generation model, requiring only a …

Standard sequential generation methods assume a pre-specified generation order, such as
text generation methods which generate words from left to right. In this work, we propose a …

The integration of algorithmic components into neural architectures has gained increased
attention recently, as it allows training neural networks with new forms of supervision such …

Recent advances in conditional recurrent language modelling have mainly focused on
network architectures (eg, attention mechanism), learning algorithms (eg, scheduled …

We present Optimal Completion Distillation (OCD), a training procedure for optimizing
sequence to sequence models based on edit distance. OCD is efficient, has no hyper …

Аннотация Приведен аналитический обзор разновидностей интегральных (end-to-end)
систем для распознавания речи, методов их построения, обучения и оптимизации …

In this paper, we explore the use of attention-based models for online speech recognition
without the usage of language models or searching. Our model is based on an attention …