A mobile device source identification method and system based on multi-mode fusion depth features

Abstract

The invention belongs to the technical field of voice forensics, and discloses a mobile device source recognition method and system based on multi-mode fusion depth features. Firstly, the test data MFCCs and GSV features are extracted, and the features are correspondingly divided into multiple channels, and then CNN is trained separately and fused. The fusion depth features are obtained, and then the trained deep residual network is used to make a judgment, and finally the judgment results of the short samples of each channel are jointly decided by the voting method. When training the GMM-UBM model, the present invention screens the data according to the characteristics of phonemes and tones of the voice data, and selects a small amount of representative data, which not only ensures the generalization of the model, but also reduces the amount of data calculation, and improves the Modeling efficiency is improved; the present invention uses a deep neural network for supervised training to extract deep features, eliminates redundancy and interference information in feature data, simplifies feature data, improves data representation, and reduces data dimension simplification amount of calculation.

Description

technical field [0001] The invention belongs to the technical field of voice forensics, and in particular relates to a mobile device source identification method and system based on multi-mode fusion depth features. Background technique [0002] Currently, the closest prior art: [0003] With the rapid development of digital media technology, various electronic products such as computers, digital cameras, mobile phones, printers, scanners, etc. have gradually become indispensable equipment in people's daily life, resulting in a large number of media files. At the same time, all kinds of professional digital media editing software are gradually becoming more convenient under the demands of people. While these editing software bring convenience and joy to people's life, they also introduce many challenging problems. Some criminals secretly recorded and forged a large amount of voice data through various recording equipment and editing software, which caused a series of probl...

AI Technical Summary

Problems solved by technology

[0007] (1) The representativeness of the characteristics of the traditional mobile device source identification method is further mined and the efficiency is poor; and the traditional judgment model is relatively intuitive, and the mobile device cannot be fully represented and modeled through the feature information; the traditional test judgment method is Based on a single decision, the recognition accuracy is low

[0008] (2) Most of the previous methods directly use the original feature data to build the algorithm model, because the original feature data has a lot of redundancy and interference information, so it increases the amount of calculation when building the algorithm model, and also makes the final The algorithm model is not accurate enough

[0009] (3) Most of the current methods use a single feature data to model the characteristics of the device source

In terms of justice, voice data is becoming more and more important as evidence, but some forged and tampered voice data conceal the truth, which brings a lot of trouble to voice recognition

Images