Skip to content
conftrace_
2021 Interspeech INTERSPEECH 2021

The LIUM Human Active Correction Platform for Speaker Diarization

Abstract

We developed a human assisted speaker diarization platform that enables a human annotator to correct the output of any speaker diarization system by providing a graphical view of the diarization segmentation and clustering steps while guiding the human annotator to optimize the correction process and easily improve the resulting diarization.

🐣 Hot Topic Early Bird - error correction

Authors

Alexandre Flucha, Anthony Larcher, Ambuj Mehrish, Sylvain Meignier, Florian Plaut, Nicolas Poupon, Yevhenii Prokopalo, Adrien Puertolas, Meysam Shamsi, Marie Tahon

Topics

Artificial Intelligence > Core AI > Human-AI Interaction Machine Learning > Learning Types > Weakly Supervised Learning

Keywords

speaker diarization error correction speaker segmentation speech annotation
Download PDF

More from Interspeech 2021

All Interspeech 2021 papers →
Beey: More Than a Speech-to-Text Editor
Downsizing of Vocal-Tract Models to Line up Variations and Reduce Manufacturing Costs
ROXANNE Research Platform: Automate Criminal Investigations
On-Device Streaming Transformer-Based End-to-End Speech Recognition
Advanced Semi-Blind Speaker Extraction and Tracking Implemented in Experimental Device with Revolving Dense Microphone Array
Forty Years of Speech and Language Processing: From Bayes Decision Rule to Deep Learning