Norwegian Speech Recognition Audio Database

This dataset was created by Nordic Language Technology for developing automatic speech recognition and dictation systems for Norwegian. The files have been renamed to be unique and meaningful, and metadata has been converted from SPL to anonymized JSON format with UTF-8 encoding. The specific number of audio files, rows, or columns is not provided in the input.

Use Cases

• Train an automatic speech recognition model on Norwegian audio files with anonymized JSON metadata.
• Develop dictation systems using the speech data originally organized for Norwegian language technology.
• Analyze the structure of renamed audio files and converted JSON metadata for ASR pipeline preprocessing.

Strengths

• Data was specifically created by Nordic Language Technology for ASR development, indicating a targeted purpose.
• File names have been made unique and meaningful to improve usability independent of folder structure.
• Metadata has been converted to JSON format and anonymized, with text encoding standardized to UTF-8.

Limitations

• The dataset size, number of audio files, and specific audio characteristics are unknown.
• The original folder structure, which contained meaningful information, has been altered, potentially losing some organizational context.
• The absence of sample data or column descriptions limits understanding of the specific metadata attributes available.

Provenance

Source

Nordic Language Technology

Collection Method

Created for the development of automatic speech recognition and dictation systems.

Time Range

null

Freshness

null

Geography

Norway