English Earnings Calls Corpus for ASR Benchmarking

A 119-hour corpus of English-language earnings calls collected from global companies. It serves as a benchmark for automatic speech recognition models on real-world accented speech.

Use Cases

• Benchmark ASR model accuracy on real-world accented speech from global companies
• Analyze acoustic and linguistic variations in English-language earnings calls
• Train speech recognition systems on a corpus of 119 hours of corporate audio

Strengths

• Corpus size of 119 hours of audio data
• Focus on real-world accented speech from global companies

Limitations

• No column or sample data available for detailed feature analysis
• Dataset size is categorized as 'n1 K', indicating a relatively small scale

Provenance

Source

huggingface

Collection Method

Collected earnings calls from global companies.

Freshness

Last updated in June 2022.

Geography

Global, with a tag indicating US region influence.