VoxEntity-EU: measured spoken-entity capture (DE/FR)

Synthetic speech corpus (automated QA only, no human listening review yet) with a simulated telephone channel; a separate human-speech control is reported in the technical report. All numbers are precalculated from frozen runs; nothing is generated or inferred here.

Language
Entity type
Condition
System
Outcome
Case (exact link: paste its id in 'Case id contains')