NOTSOFAR-1: Single-Device Far-Field Meeting Transcription Dataset is a dataset released by Microsoft in 2025 for evaluating far-field meeting transcription tasks, aiming to advance the development of single-device far-field speech separation and transcription technologies.

The dataset contains real meeting recordings covering various meeting rooms, numbers of speakers, microphone configurations, and acoustic environments. It can be used for research on far-field speech acquisition, multi-speaker overlapping speech processing, and meeting transcription. The official repository also provides baseline system code, integrating Continuous Speech Separation (CSS), Whisper-based Automatic Speech Recognition (ASR), and a NeMo-based speaker diarization module, enabling researchers to run inference pipelines and evaluate model performance.

Dataset Composition

• Dev-set: Contains single-channel and multi-channel meeting recording data. There are 177 single-channel sessions and 106 multi-channel sessions.
• Subset splits: Based on meeting scenarios and noise conditions, the data is divided into multiple subsets, including Natural Meeting (#NaturalMeeting), Debate Overlaps (#DebateOverlaps), Turns No Overlap (#TurnsNoOverlap), High Transient Noise (#TransientNoise=high), and Talking Near Whiteboard (#TalkNearWhiteboard), among others.
• Baseline system code: Provides complete Python baseline code, including continuous speech separation, Whisper-based ASR, and a NeMo-based speaker diarization module.
• Evaluation metrics: Primarily uses speaker-attributed tcpWER for ranking, with tcORC-WER serving as an auxiliary analysis metric.
• Training data: In addition to the dev-set, a Simulated Training Set is provided, containing approximately 1,000 hours or 200 hours of simulated data for model training and fine-tuning.