Výsledky vyhledávání

Report

Useful Blunders: Can Automated Speech Recognition Errors Improve Downstream Dementia Classification?

Autor: Li, Changye, Xu, Weizhe, Cohen, Trevor, Pakhomov, Serguei

\textbf{Objectives}: We aimed to investigate how errors from automatic speech recognition (ASR) systems affect dementia classification accuracy, specifically in the ``Cookie Theft'' picture description task. We aimed to assess whether imperfect ASR-g

Externí odkaz: http://arxiv.org/abs/2401.05551

Zobrazit plný text záznamu

Report

TRESTLE: Toolkit for Reproducible Execution of Speech, Text and Language Experiments

Autor: Li, Changye, Xu, Weizhe, Cohen, Trevor, Michalowski, Martin, Pakhomov, Serguei

The evidence is growing that machine and deep learning methods can learn the subtle differences between the language produced by people with various forms of cognitive impairment such as dementia and cognitively healthy individuals. Valuable public d

Externí odkaz: http://arxiv.org/abs/2302.07322

Zobrazit plný text záznamu

Report

GPT-D: Inducing Dementia-related Linguistic Anomalies by Deliberate Degradation of Artificial Neural Language Models

Autor: Li, Changye, Knopman, David, Xu, Weizhe, Cohen, Trevor, Pakhomov, Serguei

Deep learning (DL) techniques involving fine-tuning large numbers of model parameters have delivered impressive performance on the task of discriminating between language produced by cognitively healthy individuals, and those with Alzheimer's disease

Externí odkaz: http://arxiv.org/abs/2203.13397

Zobrazit plný text záznamu