Advancements in Audio Tampering and Forgery Detection Techniques

Audio has become a crucial part of communication, legal proceedings, digital documentation etc. Maintaining authenticity of audio recordings is an acute concern in this digital age. Audio forensics deals with acquisition, analysis and authentication of audio recordings. This plays a critical part in evidence presented in legal or investigative contexts. Detecting tampered audio has become extremely difficult with the availability of advanced audio editing tools and deep generative models. This literature review scrutinizes the advancement and current status of the techniques developed for detecting and localizing tampered audio. It deals with an extensive analysis of techniques based on Electrical Network Frequency (ENF), environmental and device-specific signatures, and signal processing to determine various types of tampering. Various machine learning and deep learning techniques are also used to assess audio signals. They detect defects in time-domain, frequency-domain features and temporal patterns present. Metadata and compression artifact analysis examines anomalies in audio file structure, codec parameters and compression histories. The review lays emphasis on the strengths and weaknesses of existing frameworks in audio forensics. This work encourages development of effective and reliable techniques for audio manipulation detection.