# The role of context in neural pitch accent detection in English

arXiv: https://arxiv.org/abs/2004.14846 · HTML: https://mlsanity.com/paper/2004.14846/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 30 Apr 2020.

## Built on by (most cited first, 5 of 5)

- [Automatic recognition of suprasegmentals in speech](https://mlsanity.com/paper/2108.01122/) (2021)
- [Deep Learning For Prominence Detection In Children's Read Speech](https://mlsanity.com/paper/2110.14273/) (2021)
- [Towards Spontaneous Style Modeling with Semi-supervised Pre-training for Conversational Text-to-Speech Synthesis](https://mlsanity.com/paper/2308.16593/) (2023)
- [Integrating Feedback Loss from Bi-modal Sarcasm Detector for Sarcastic Speech Synthesis](https://mlsanity.com/paper/2508.13028/) (2025)
- [CNN Encoding of Acoustic Parameters for Prominence Detection](https://mlsanity.com/paper/2104.05488/) (2021)

## Abstract (arXiv)

Prosody is a rich information source in natural language, serving as a marker for phenomena such as contrast. In order to make this information available to downstream tasks, we need a way to detect prosodic events in speech. We propose a new model for pitch accent detection, inspired by the work of Stehwien et al. (2018), who presented a CNN-based model for this task. Our model makes greater use of context by using full utterances as input and adding an LSTM layer. We find that these innovations lead to an improvement from 87.5% to 88.7% accuracy on pitch accent detection on American English speech in the Boston University Radio News Corpus, a state-of-the-art result. We also find that a simple baseline that just predicts a pitch accent on every content word yields 82.2% accuracy, and we suggest that this is the appropriate baseline for this task. Finally, we conduct ablation tests that show pitch is the most important acoustic feature for this task and this corpus.