TAU Urban Acoustic Scenes 2019 Mobile: 10-Second Audio Segments from 10 Urban Environments

30 hours of audio data collected by Toni Heittola at Tampere University. The dataset consists of 10-second audio segments recorded in mobile settings across 10 distinct urban acoustic scenes, such as airports, metro stations, and parks. It serves as an evaluation dataset for acoustic scene classification tasks.

Use Cases

• Training acoustic scene classifiers based on the 10 labeled urban environments.
• Benchmarking audio classification models using the standardized 10-second segments.
• Studying urban soundscapes and noise pollution patterns based on the recorded scenes.
• Developing mobile audio recognition systems based on the 'mobile' recording context.

Strengths

• Contains 30 hours of audio, providing a substantial amount of data for model training.
• Focuses on 10 distinct and common urban acoustic scenes, offering clear classification targets.
• Audio segments are standardized to 10 seconds, which likely simplifies preprocessing and model input.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.
• The description does not specify recording locations or device details, which may limit reproducibility.

Provenance

Source

Tampere University

Collection Method

Likely contains audio recordings gathered in mobile settings across various urban locations.

Time Range

2019

Freshness

Last updated date is unknown.

Geography

Urban environments; specific locations are not stated.