Skip to content

TeachFX 300-hour manually transcribed classroom audio dataset

Product or Programme · All products and programmes
Evidence · 1 claim (1 for) · 1 study (1 design), q2 · 0 of 1 report an effect size · 1 claim rests on one study

Description

A 300-hour manually transcribed classroom-audio dataset created by TeachFX to improve transcription models and support research on equitable automated speech recognition across diverse English dialects.

Components

  • TeachFX 300-hour manually transcribed classroom audio dataset for equitable ASR research: TeachFX manually transcribed 300 hours of classroom audio recordings to create a dataset with dual purposes: refining transcription models within TeachFX to improve product accuracy, and providing external researchers with resources for more equitable AI and education research. The project aimed to enhance transcription for teachers and students who speak diverse cultural and regional English dialects, addressing known biases in automated speech recognition. (An Ethical and Equitable Vision of AI in Education: Learning Across 28 Exploratory Project)

Claims

-

Key Sources

  • Noakes, S., Shell, A., Murillo, A., Van Nostrand, P., Ruiz, P., Cornick, S., and Karim, S. (2024, October). An Ethical and Equitable Vision of AI in Education: Learning Across 28 Exploratory Projects. Digital Promise. https://doi.org/10.51388/20.500.12265/232
Page metadata
Field Value
type product
id teachfx-300-hour-manually-transcribed-classroom-audio-dataset
title TeachFX 300-hour manually transcribed classroom audio dataset
description A 300-hour manually transcribed classroom-audio dataset created by TeachFX to improve transcription models and support research on equitable automated speech recognition across diverse English dialects.
status review
moved_from element/teachfx-300-hour-classroom-audio-dataset
product_kind dataset
generated by process:sweep-kinds on 2026-10-10
trust unverified — no one has confirmed this page's content
sources 1 — noakes-s-shell-a-murillo

Frontmatter is the page's machine-readable record — see the Schema & Guide.