YoVDO

Spoken Corpora Data, Automatic Speech Recognition, and Bias Against African American Language - The Case of Habitual 'Be'

Offered By: Association for Computing Machinery (ACM) via YouTube

Tags

ACM FAccT Conference Courses

Course Description

Overview

Explore a research presentation examining the biases in automatic speech recognition systems against African American Language, focusing on the case of habitual 'be'. Delve into the analysis of spoken corpora data and its implications for natural language processing technologies. Learn about the challenges faced in accurately recognizing and processing diverse linguistic features, particularly those present in African American English. Gain insights into the importance of addressing these biases to create more inclusive and equitable language technologies.

Syllabus

Spoken Corpora Data, Automatic Speech Recognition, and Bias Against African American Language: The C


Taught by

ACM FAccT Conference

Related Courses

Translation Tutorial - Thinking Through and Writing About Research Ethics Beyond "Broader Impact"
Association for Computing Machinery (ACM) via YouTube
Translation Tutorial - Data Externalities
Association for Computing Machinery (ACM) via YouTube
Translation Tutorial - Causal Fairness Analysis
Association for Computing Machinery (ACM) via YouTube
Implications Tutorial - Using Harms and Benefits to Ground Practical AI Fairness Assessments
Association for Computing Machinery (ACM) via YouTube
Responsible AI in Industry - Lessons Learned in Practice
Association for Computing Machinery (ACM) via YouTube