Speech Translation And Summarization: English-Centric Multilingual Audio Dataset

McGill-NLP provides generated article and summary audio for English-centric multilingual directions. The dataset includes audio files and metadata for language pairs such as Amharic-English, Arabic-English, Bengali-English, and Chinese-English. It was last updated on June 17, 2026.

Use Cases

• Training speech translation models based on the described English-centric multilingual audio pairs.
• Evaluating summarization systems using the generated summary audio files mentioned in the description.
• Benchmarking few-shot learning performance for audio tasks using the described few_shot splits.
• Studying cross-lingual transfer in speech processing based on the included language directions.

Strengths

• Includes audio and metadata for multiple language pairs, such as Amharic-English and Arabic-English.
• Organized into structured splits, including few_shot and test sets as mentioned in the description.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Freshness should be verified as the last update is from 2026.

Provenance

Source

McGill-NLP

Collection Method

Dataset contains generated article and summary audio.

Freshness

Last updated 2026-06-17 08:13:08