Fact · Transcription
20percent relative DER reduction
On the AMI meeting corpus, assigning overlapped speech with overlap-aware resegmentation reduced the diarization error rate of the baseline system by 20 percent relative to that baseline.
Checked against its source on October 17, 2026
In the source's own words
“We apply overlap-aware resegmentation on AMI, resulting in a 20% relative DER reduction over the baseline system.”
Overlap-aware diarization: resegmentation using neural end-to-end overlapped speech detection, Latané Bullock et al., arXiv (submitted to ICASSP 2020), 2019, Abstract.
Open the original- What it measures
- AMI is a corpus of recorded meetings. The method first detects overlapped speech with a neural model, then gives each overlapped region two speaker labels.
- What it does not show
- The figure is a relative reduction, and the abstract does not give the baseline's absolute error. Overlap gains were modest in the authors' own follow-up work.