Voice deformation detection method based on deep dense network
A technology of dense network and detection method, applied in the field of information security, can solve the problem that the anti-attack capability of ASV system needs to be further improved, and the confusion of true and false voices.
Pending Publication Date: 2021-07-06
GUANGDONG POLYTECHNIC NORMAL UNIV
View PDF0 Cites 2 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Vulnerabilities that easily confuse real and fake speech remain largely unknown, although several automated speaker verification research groups are proposing countermeasures
There are still some loopholes in the speaker verification system for different spoofing attacks, and the anti-attack capability of the ASV system needs to be further improved
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.
[0020] refer to Figure 1~3 , the present invention is a kind of voice deformation detection method based on depth dense network, mainly comprises the following steps:
[0021] First, build a deep dense learning network structure based on convolutional neural network. Each convolutional layer of the network structure will accept all previous convolutional layers as its additional input, and the input of each convolutional layer is its The output of the previo...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention relates to a voice deformation detection method based on a deep dense network. The method is characterized by comprising the steps of firstly constructing a deep dense learning network structure based on a convolutional neural network, then training the cross entropy error of the network structure through small-batch iteration stochastic gradient descent, and finally, inputting the time-frequency characteristics of the to-be-detected voice into a trained deep dense network structure, judging the authenticity of the to-be-detected voice by a softmax layer in the network structure through a camouflage factor, identifying a camouflage tool correspondingly used by the camouflage voice, and outputting a detection result. According to the method, whether the to-be-detected voice is the original voice or the voice subjected to deformation camouflage can be identified, and a tool for further detecting deformation camouflage can be further detected while the deformed voice can be classified.
Description
technical field [0001] The present invention relates to the technical field of information security, and more specifically, the present invention relates to a voice deformation detection method based on a deep dense network. Background technique [0002] With the development of computer and other information technology. Many audio processing software programs, such as Audacity, CoolEdit, PRAAT, Real-Time Iterative Spectrogram (RTIS) algorithm based on matlab tools, and voice deformation, are widely used in audio forensics, entertainment, privacy protection and other fields. Misuse of these software products has also increased voice-based illegal activities such as online fraud, voice payments, phone chats, etc. It is easy for people to disguise their own recordings as other voices on their personal computers or smartphones and send them out, so that the speaker's voice changes in real time and then transmitted, so that the receiving party cannot recognize the speaker's iden...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G10L25/51G10L25/30G10L25/18G06N3/04G06N3/08G06K9/62
CPCG10L25/51G10L25/30G10L25/18G06N3/084G06N3/045G06F18/253
Inventor 王泳张奥运
Owner GUANGDONG POLYTECHNIC NORMAL UNIV