Graph2Speak: Improving Speaker Identification using Network Knowledge in Criminal Conversational Data
This work addresses speaker identification for criminal investigations, but it is incremental as it builds on existing methods with small gains.
The paper tackled speaker identification in criminal conversational data by re-ranking candidates based on interaction frequency, improving baseline accuracy by 1.2% absolute on CSI data and 1.1% on ROXANNE data.
Criminal investigations mostly rely on the collection of speech conversational data in order to identify speakers and build or enrich an existing criminal network. Social network analysis tools are then applied to identify the most central characters and the different communities within the network. We introduce two candidate datasets for criminal conversational data, Crime Scene Investigation (CSI), a television show, and the ROXANNE simulated data. We also introduce the metric of conversation accuracy in the context of criminal investigations. By re-ranking candidate speakers based on the frequency of previous interactions, we improve the speaker identification baseline by 1.2% absolute (1.3% relative), and the conversation accuracy by 2.6% absolute (3.4% relative) on CSI data, and by 1.1% absolute (1.2% relative), and 2% absolute (2.5% relative) respectively on the ROXANNE simulated data.