Error Analysis in a Modular Meeting Transcription System
This work addresses error analysis for meeting transcription systems, providing incremental insights into leakage and segmentation effects.
The paper tackled error analysis in meeting transcription by extending a leakage analysis framework for speech separation, revealing significant cross-channel leakage during single-speaker activity that minimally impacts final performance due to VAD filtering, and showing advanced diarization reduces the gap to oracle segmentation by a third compared to simple VAD.
Meeting transcription is a field of high relevance and remarkable progress in recent years. Still, challenges remain that limit its performance. In this work, we extend a previously proposed framework for analyzing leakage in speech separation with proper sensitivity to temporal locality. We show that there is significant leakage to the cross channel in areas where only the primary speaker is active. At the same time, the results demonstrate that this does not affect the final performance much as these leaked parts are largely ignored by the voice activity detection (VAD). Furthermore, different segmentations are compared showing that advanced diarization approaches are able to reduce the gap to oracle segmentation by a third compared to a simple energy-based VAD. We additionally reveal what factors contribute to the remaining difference. The results represent state-of-the-art performance on LibriCSS among systems that train the recognition module on LibriSpeech data only.