# Comparing Supervised Models And Learned Speech Representations For Classifying Intelligibility Of Disordered Speech On Selected Phrases

arXiv: https://arxiv.org/abs/2107.03985 · HTML: https://mlsanity.com/paper/2107.03985/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 35 references that are in the map.
- First posted on arXiv 8 Jul 2021.

## Builds on (most cited first, 12 of 12)

- [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://mlsanity.com/paper/1704.04861/) (2017)
- [UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction](https://mlsanity.com/paper/1802.03426/) (2018)
- [CNN Architectures for Large-Scale Audio Classification](https://mlsanity.com/paper/1609.09430/) (2016)
- [Speech Recognition with Deep Recurrent Neural Networks](https://mlsanity.com/paper/1303.5778/) (2013)
- [Streaming End-to-end Speech Recognition For Mobile Devices](https://mlsanity.com/paper/1811.06621/) (2018)
- [Towards Learning a Universal Non-Semantic Representation of Speech](https://mlsanity.com/paper/2002.12764/) (2020)
- [Recognizing long-form speech using streaming end-to-end models](https://mlsanity.com/paper/1910.11455/) (2019)
- [LEAF: A Learnable Frontend for Audio Classification](https://mlsanity.com/paper/2101.08596/) (2021)
- [Personalizing ASR for Dysarthric and Accented Speech with Limited Data](https://mlsanity.com/paper/1907.13511/) (2019)
- [Learning to detect dysarthria from raw speech](https://mlsanity.com/paper/1811.11101/) (2018)
- [Automatic dysarthric speech detection exploiting pairwise distance-based convolutional neural networks](https://mlsanity.com/paper/2011.07545/) (2020)
- [Automatic and perceptual discrimination between dysarthria, apraxia of speech, and neurotypical speech](https://mlsanity.com/paper/2011.07542/) (2020)

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

- [BigSSL: Exploring the Frontier of Large-Scale Semi-Supervised Learning for Automatic Speech Recognition](https://mlsanity.com/paper/2109.13226/) (2021)
- [Universal Paralinguistic Speech Representations Using Self-Supervised Conformers](https://mlsanity.com/paper/2110.04621/) (2021)
- [TRILLsson: Distilled Universal Paralinguistic Speech Representations](https://mlsanity.com/paper/2203.00236/) (2022)
- [Clinical BERTScore: An Improved Measure of Automatic Speech Recognition Performance in Clinical Settings](https://mlsanity.com/paper/2303.05737/) (2023)
- [Self-supervised Neural Factor Analysis for Disentangling Utterance-level Speech Representations](https://mlsanity.com/paper/2305.08099/) (2023)

## Abstract (arXiv)

Automatic classification of disordered speech can provide an objective tool for identifying the presence and severity of speech impairment. Classification approaches can also help identify hard-to-recognize speech samples to teach ASR systems about the variable manifestations of impaired speech. Here, we develop and compare different deep learning techniques to classify the intelligibility of disordered speech on selected phrases. We collected samples from a diverse set of 661 speakers with a variety of self-reported disorders speaking 29 words or phrases, which were rated by speech-language pathologists for their overall intelligibility using a five-point Likert scale. We then evaluated classifiers developed using 3 approaches: (1) a convolutional neural network (CNN) trained for the task, (2) classifiers trained on non-semantic speech representations from CNNs that used an unsupervised objective [1], and (3) classifiers trained on the acoustic (encoder) embeddings from an ASR system trained on typical speech [2]. We found that the ASR encoder's embeddings considerably outperform the other two on detecting and classifying disordered speech. Further analysis shows that the ASR embeddings cluster speech by the spoken phrase, while the non-semantic embeddings cluster speech by speaker. Also, longer phrases are more indicative of intelligibility deficits than single words.