kzhou/sf_streets

Dataset for paper: "Sorry, I Didn't Catch That": How Speech Models Miss What Matters Most

https://arxiv.org/abs/2602.12249

Note: This dataset is a public version of the dataset used in the paper due to participant privacy constraints.

Forbidden Uses:

• Generating, enabling, or promoting hate speech, harassment, discrimination, misinformation, or culturally offensive or harmful content
• Voice cloning, speaker impersonation, or the creation of synthetic voices intended to resemble or replicate any participant
• Attempting to identify, re-identify, or infer the identity of any participant, including attempts to extract personally identifiable information from the audio or associated metadata
• Any commercial, for-profit, or revenue-generating use, including product development, advertising, or monetized services
• Any use that misrepresents, stereotypes, or falsely attributes characteristics, language abilities, accents, or identities to the speakers
• Redistribution of the dataset under terms that conflict with or weaken these restrictions

Participants were recruited via Prolific and paid $15 USD an hour for their contributions. IRB was reviewed by NorthStar IRB and participants gave consent to have their anonymized audio clips be publicly shared for research.