No Need for a Lexicon? Evaluating the Value of the Pronunciation Lexica in End-to-End Models
By Tara N. Sainath, Rohit Prabhavalkar, Shankar Kumar, Seungjin Lee, Anjuli Kannan, David Rybach, Vlad Schogol, Patrick Nguyen, Bo Li, Yonghui Wu and 2 more. 2017
54 citations recorded by the source corpus.
For decades, context-dependent phonemes have been the dominant sub-word unit for conventional acoustic modeling systems. This status quo has begun to be challenged recently by end-to-end models which seek to combine acoustic, pronunciation, and language model components into a single neural network. Such systems, which typically predict graphemes or words, simplify the recognition process since they remove the need for a separate expert-curated pronunciation lexicon to map from phoneme-based units to words. However, there has been little previous work comparing phoneme-based versus grapheme-based sub-word units in the end-to-end modeling framework, to determine whether the gains from such approaches are primarily due to the new probabilistic model, or from the joint learning of the various components with grapheme-based units. In this work, we conduct detailed experiments which are aimed at quantifying the value of phoneme-based pronunciation lexica in the context of end-to-end models. We examine phoneme-based end-to-end models, which are contrasted against grapheme-based ones on a large vocabulary English Voice-search task, where we find that graphemes do indeed outperform phonemes. We also compare grapheme and phoneme-based approaches on a multi-dialect English task, which once again confirm the superiority of graphemes, greatly simplifying the system for recognizing multiple dialects.
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
Citations in the corpus
- A Comparison of Modeling Units in Sequence-to-Sequence Speech Recognition with the Transformer on Mandarin Chinese (2018, background): “A S] 1 8 M ay 2 01 8 thermore, the modeling units of WPM have been explored in [6], which are sub-word units ranging from graphemes all the way up to entire words.”
- Model Unit Exploration for Sequence-to-Sequence Speech Recognition (2019, methodology): “5As also reported in [17]; though, we note that we get about 2% absolute degradation in WERs with a model trained without EOW.”
- Advancing Multi-Accented Lstm-CTC Speech Recognition Using a Domain Specific Student-Teacher Learning Paradigm (2018, methodology): “Given training data of multiple conditions, we need to exploit the data efficiently to train an improved multi-condition (domain) ASR model [5, 6, 7, 8].”
- From Senones to Chenones: Tied Context-Dependent Graphemes for Hybrid Speech Recognition (2019, background): “, “t h e c a t”) [17,23,35], word piece level (e.”
- Hybrid Autoregressive Transducer (HAT) (2020, background): “This has led some to conclude a lexicon or external LM is unnecessary [23, 24, 25].”
- Recent Advances in End-to-End Automatic Speech Recognition (2021, background): “While most units (characters [74, 80], word-piece [37, 101, 244, 308, 322], words [7, 154, 259], or even phrases [67]) are purely derived based on text and word-piece unit is the dominating one, there are also studies working on phoneme units [112, 234, 293, 345].”
- State-of-the-Art Speech Recognition with Sequence-to-Sequence Models (2017, background): “One of the disadvantages of using phonemes is that it requires having an additional PM and LM, and was not found to improve over graphemes in our experiments [19].”
- Lingvo: a Modular and Scalable Framework for Sequence-to-Sequence Modeling (2019, background): “To date, this framework has produced a number of state-of-the-art results in machine translation [2, 21, 23], speech recognition [3, 4, 5, 10, 11, 12, 13, 14, 16, 15, 20, 22], speech synthesis [6, 8, 19], and speech translation [7, 9].”