Developing a secure voice recognition service on Raspberry Pi (original) (raw)
Related papers
Bio-metric Encryption of Data Using Voice Recognition
Automation, Control and Intelligent Systems, 2021
In symmetric cryptosystems, the protection of secret keys is based on the traditional user authentication and likewise the security of the cryptosystem depends on the secrecy of the secret keys. In the event of lost, theft or infection of these secrete keys; the security of the cryptosystems would be compromised hence exposing critical information. Biometrics has been commercially used to verify user's identity. Voice biometrics has been proven to be even more effective because it cannot be stolen in some cases like face, fingerprint or even iris biometrics. The research proves that a well-designed system will prompt an authentication question and on verification user must provide both the desired answer as well as desired matching threshold or the system ignores the user features. This research proposes a software-based architecture solution for Biometric Encryption of data using Voice Recognition that employed the Dynamic Time Warping (DTW) technique to solve the problem of speech biometric duration varying with non-linear expansion and contraction. The approach then used database to store the monolithically bind cryptographic key with the equivalent biometric hardened template of the user in such manner that identity of the key will stay hidden unless there is a successful biometric authentication by intended party. The research used the MIT mobile device speaker verification corpus (MDB) and A data set in quiet environment (QDB) for training and verifying session. Finally using the Equal Error Rate (EER) the research evaluated performance or rate at which False Acceptance Rate (FAR) and a False Rejection Rate (FRR) are equal. Therefore, according to the result it offers a better substitute method of user authentication than traditional pre-shared keys for benefit of protecting secret keys.
With increasing technology development in field of communication and Electronic devices, there is a need for better security service for information transfer in Medical Sectors, Banking, Financial and in other IoT applications etc. Fight against security attacks is of prime importance. Through Cryptographic techniques we can provide Authenticity as well as Confidentiality for the user data. In this paper, hardware implementation has been described for a real-time application of speech data encryption and decryption using AES algorithm along with the speech recognition using cross correlation technique. Verilog programming environment is used for AES cryptography whereas MATLAB is used for Speech recognition. ASIC design on AES core is implemented using Cadence tools. Number of gates, area and power used by AES core design has been drastically reduced by specifying wide range of constraints during front end designing. In Backend designing, layout of AES design, which is the physical geometric representation is also developed.
REVIEW: SECURE AND RELIABLE VOICE RECOGNITION SYSTEM USING DEEP NEURAL NETWORKS ACOUSTIC MODEL
IAEME PUBLICATION, 2021
In recent days, voice recognition system is used in various impending segments like banking, communication, industrial and government sectors. Most applications are used in real time which are vulnerable to various security threats. It causes big contests on various levels in data protection to tackle security threats in cyber space. Performance of secure voice recognition system is measured. Accuracy is evaluated with word Error Rate (WER) and speed is restrained with factors such as vocabulary size, language constraints, confusability speech and image recognition will be recorded at each stage processing will be monitored. Confrontation of these types of known or unknown risks, several privacy-preserving approaches have been proposed and used for either Secure Multi-party Computation (SMC) or homomorphic encryption (HE) as the underlying schemes. In last decade, technology have been developed and voice recognition have come a long way. From magic eight ball ancient world to today’s modern technology voice recognition system is the major cause. Tech giants in the industry were seeing the intense level of competition and the companies jumping to create content in the space because of increasing prevalence.
Achieving Security Via Speech Recognition
International Journal of Information Sciences and Techniques, 2016
Speech is one of the essential sources of the conversation between human beings. We as humans speak and listen to each other in human-human interface. People have tried to develop systems that can listen and prepare a speech as persons do so naturally. This paper presents a brief survey on Speech recognition, allow people to compose documents and control their computers with their voice. In other words, the process of enabling a machine (like a computer) to identify and respond to the sounds produced in human speech. ASR can be treated as the independent, computer-driven script of spoken language into readable text in real time. The Speech Recognition system requires careful attention to the following issues: Meaning of various types of speeches, speech representation, feature extraction techniques, speech classifiers, and database and performance evaluation. This paper helps in understanding the technique along with their pros and cons. A comparative study of different technique is done as per stages.
A SIMPLE METHOD TO ENCRYPT-DECRYPT SPEECH SIGNAL
Speech signal some time contains secret information, and to prevent unauthorized party to understand the speech signal we will introduce a simple method of speech signal encryption-decryption. The proposed method will use a huge color image as a private key, this key will not be transmitted and it will be kept by the sender and the receiver. The proposed method will be tested and implemented to prove the efficiency of the proposed method. The quality parameters will be calculated to insure that the proposed method will meet the requirements of good method of encryption-decryption.
Simple, Qualities, Efficient and Secure Method to Encrypt Voice Signal
International Journal of Computer Applications, 2021
Voice signal cryptography is needed process to protect the voice signal from any third party, because the wave file may be confidential or may contain valuable secret information. In tis paper research we will introduce a method of voice signal encryption-decryption base on using a huge RGB color image as secret key, this image cab be easily used to create encryption-decryption private key. It will be shown that the proposed method will be qualities by destroying the original file to generate the encrypting one, and recovering identical to the original decrypted file. The efficiency the proposed method will be calculated to show the productivity of this method.
Secured Audio Encryption using AES Algorithm
International Journal of Computer Applications
Speech to text conversion is the process of converting spoken words into written texts. In this paper, a voice encryption system is developed as a real-time software application. Basically, the speech is taken as an input and is encoded to be decoded by authenticated users only. The algorithm used to perform this cryptography is Advanced Encryption Standards (AES). This algorithm has its own particular structure to encrypt and decrypt sensitive data and is applied in hardware and software all over the world. We take the information in the form of audio as an input using a microphone which is to be transmitted over the channel to the intended receiver. The audio input is converted into text format which ensures speech-to-text conversion. Then, the text is encrypted using the AES algorithm to form cipher text. This cipher text is sent over a channel to the receiver. The receiver requests to perform decryption of the information only if he has the correct secret key otherwise the request is declined. If the key matches, the decryption is successful and the receiver get the message as text. For 128 bit, about 2128 attempts are needed to break. This makes it very difficult to hack it as a result it is very safe protocol.
Implementation of RSA Algorithm for Speech Data Encryption and Decryption
Today, organizations in both public and private sectors have become increasingly dependent on electronic data processing. This digital data are going through an insecure channel from one place to another and anyone can easily get those important data without the concerns of the sender. So, protecting these important data is crucial task in data communication and Public Key Cryptography is one of the best ways to protect digital data from the unauthorized access. RSA algorithm is extensively used in the popular implementations of Public Key Infrastructures. In this paper, we have done an efficient implementation of RSA algorithm for speech data encryption and decryption. At first, five hundred Bangla speech words were recorded from six different speaker and stored as RIFF (.wav) file format. Then our developed program was used to extract data from these words and this data were stored in a text file as integer data. Finally, we used our implemented program to encrypt and decrypt speech data.
Design and Hardware Implementation of a Speech Cipher System
Al-Nahrain Journal for Engineering Sciences, 2007
Digital ciphering of speech signals based on one of modern cryptography algorithms, called the Rijndael algorithm, is studied and presented in this paper. The algorithm meets most of the requirements of security level in recent applications. A system to encrypt speech files recorded with Sound Blaster Card of a personal computer was proposed and simulated successfully using MATLAB® language. Subjective measure and objective measure using segmental spectral signal-to-noise ratio, were used to test the proposed system performance. In these tests residual intelligibility of the encrypted speech and quality of the recovered speech were calculated and assessed. Finally, a hardware implementation of the above cipher system has been proposed using the TMS320-C30. The real time requirements from the speech cipher system have been computed in terms of execution time together with factors affecting such implementation. The results show the capability of the cipher system to be implemented usi...