Forensic Audio Classifier Tool is an ML-based digital forensics system built using PyTorch, Transformers, and a custom hybrid pipeline (Acoustic Model + Language Model + Classifier). It is designed for the Tripura Bengali dialect, enabling accurate transcription, keyword detection, and automated (Flagged / Review / Safe) audio classification.
-
Updated
Nov 14, 2025 - Python