SecureLogix Audio Analysis System

offline

Inference playground

Synthetic-speech analysis · calibrated flagship (12.91% ITW) + two-model fusion (9.97% CV)

checking…
  1. 1Load target
  2. 2Run detection
  3. 3Review verdict

Point the detector at a suspect clip — drop an audio file (or pick one via “Try a sample”), then run detection for a calibrated real-vs-synthetic verdict. Calibrated on In-the-Wild; ~12.9% EER out-of-domain — accuracy drops on audio unlike the training corpus.

Detection targetload a clip, then run detection
converted to 16-bit WAV locally — loads as the target, never auto-scores