On Training Recurrent Networks With Truncated Backpropagation Through Time In Speech Recognition

Abstract

Recurrent neural networks have been the dominant models for many speech and language processing tasks. However, we understand little about the behavior and the class of functions recurrent networks can realize. Moreover, the heuristics used during training complicate the analyses. In this paper, we study recurrent networks' ability to learn long-term dependency in the context of speech recognition. We consider two decoding approaches, online and batch decoding, and show the classes of functions to which the decoding approaches correspond. We then draw a connection between batch decoding and a popular training approach for recurrent networks, truncated backpropagation through time. Changing the decoding approach restricts the amount of past history recurrent networks can use for prediction, allowing us to analyze their ability to remember. Empirically, we utilize long-term dependency in subphonetic states, phonemes, and words, and show how the design decisions, such as the decoding appr

Stats

• citations12
• S2 citations—
• github stars0
• HF likes0
• heat score8.35
• arxiv keytang2018on

Related papers

• Exploring Architectures, Data And Units For Streaming End-to-end Speech Recognition With Rnn-transducer (2018)16.21
• Memory Visualization For Gated Recurrent Neural Networks In Speech Recognition (2016)11.76
• Improving RNN Transducer Modeling For End-to-end Speech Recognition (2019)0.00
• A Comparison Of End-to-end Models For Long-form Speech Recognition (2019)12.93
• Learning The Sequential Temporal Information With Recurrent Neural Networks (2018)0.00
• Segmental Recurrent Neural Networks For End-to-end Speech Recognition (2016)0.00
• Frame Stacking And Retaining For Recurrent Neural Network Acoustic Model (2017)0.00
• Improving Speech Recognition By Revising Gated Recurrent Units (2017)11.19