An Efficient Time-Frequency Domain Speech Perceptual Hashing Authentication Algorithm Based on Discrete Wavelet Transform

Abstract

According to the situation that speech authentication algorithms are not appropriated for real-time speech content authentication, a novel speech perceptual hashing authentication algorithm based on discrete wavelet transform with combination of time-frequency domain features was proposed. Firstly, by discrete wavelet transform (DWT), a new signal in frequency domain is generated from the original speech signal after pre-processing and intensity-loudness transform (ILT). Next, the algorithm partitions low frequency wavelet decomposition coefficients into equal-sized and nonoverlapping blocks, and then computes logarithmic short-time energy of each block to obtain speech signal's features in frequency domain. Finally, combining with spectral flux features (SFF) of speech signal in time domain, a ternary perceptual hashing sequence is created. Experiment results show that ternary form is better to stand for hashing digest than binary form; the proposed algorithm has a good robustness against content preserving operations, discrimination, compactness and high efficiency, and detects the tamper localization as well.