# SpeechBlender: Speech Augmentation Framework for Mispronunciation Data Generation

arXiv: https://arxiv.org/abs/2211.00923 · HTML: https://mlsanity.com/paper/2211.00923/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 6 of its 28 references that are in the map.
- First posted on arXiv 2 Nov 2022.

## Builds on (most cited first, 6 of 6)

- [speechocean762: An Open-Source Non-native English Speech Corpus For Pronunciation Assessment](https://mlsanity.com/paper/2104.01378/) (2021)
- [Transformer-Based Multi-Aspect Multi-Granularity Non-Native English Speaker Pronunciation Assessment](https://mlsanity.com/paper/2205.03432/) (2022)
- [Context-aware Goodness of Pronunciation for Computer-Assisted Pronunciation Training](https://mlsanity.com/paper/2008.08647/) (2020)
- [3M: An Effective Multi-view, Multi-granularity, and Multi-aspect Modeling Approach to English Pronunciation Assessment](https://mlsanity.com/paper/2208.09110/) (2022)
- [A Full Text-Dependent End to End Mispronunciation Detection and Diagnosis with Easy Data Augmentation Techniques](https://mlsanity.com/paper/2104.08428/) (2021)
- [Improving Non-native Word-level Pronunciation Scoring with Phone-level Mixup Data Augmentation and Multi-source Information](https://mlsanity.com/paper/2203.01826/) (2022)

## Built on by (most cited first, 2 of 2)

- [L1-aware Multilingual Mispronunciation Detection Framework](https://mlsanity.com/paper/2309.07719/) (2023)
- [Light-weight Pronunciation Assessment via Discrete Speech Token Surprisal](https://mlsanity.com/paper/2606.19910/) (2026)

## Abstract (arXiv)

The lack of labeled second language (L2) speech data is a major challenge in designing mispronunciation detection models. We introduce SpeechBlender - a fine-grained data augmentation pipeline for generating mispronunciation errors to overcome such data scarcity. The SpeechBlender utilizes varieties of masks to target different regions of phonetic units, and use the mixing factors to linearly interpolate raw speech signals while augmenting pronunciation. The masks facilitate smooth blending of the signals, generating more effective samples than the `Cut/Paste' method. Our proposed technique achieves state-of-the-art results, with Speechocean762, on ASR dependent mispronunciation detection models at phoneme level, with a 2.0% gain in Pearson Correlation Coefficient (PCC) compared to the previous state-of-the-art [1]. Additionally, we demonstrate a 5.0% improvement at the phoneme level compared to our baseline. We also observed a 4.6% increase in F1-score with Arabic AraVoiceL2 testset.