Privacy-Preserving Machine Learning for Speech Processing [electronic resource] / by Manas A. Pathak.Material type: TextLanguage: English Series: Springer Theses, Recognizing Outstanding Ph.D. Research: Publisher: New York, NY : Springer New York : Imprint: Springer, 2013Description: XVII, 141 p. 21 illus., 13 illus. in color. online resourceContent type: text Media type: computer Carrier type: online resourceISBN: 9781461446392Subject(s): Engineering | Data structures (Computer science) | Telecommunication | Production of electric energy or power | Engineering | Signal, Image and Speech Processing | Communications Engineering, Networks | Data Structures, Cryptology and Information Theory | Power Electronics, Electrical Machines and NetworksAdditional physical formats: Printed edition:: No titleDDC classification: 621.382 LOC classification: TK5102.9TA1637-1638TK7882.S65Online resources: Click here to access online
Thesis Overview -- Speech Processing Background -- Privacy Background -- Overview of Speaker Verification with Privacy -- Privacy-Preserving Speaker Verification Using Gaussian Mixture Models -- Privacy-Preserving Speaker Verification as String Comparison -- Overview of Speaker Indentification with Privacy -- Privacy-Preserving Speaker Identification Using Gausian Mixture Models -- Privacy-Preserving Speaker Identification as String Comparison -- Overview of Speech Recognition with Privacy -- Privacy-Preserving Isolated-Word Recognition -- Thesis Conclusion -- Future Work -- Differentially Private Gaussian Mixture Models.
This thesis discusses the privacy issues in speech-based applications, including biometric authentication, surveillance, and external speech processing services. Manas A. Pathak presents solutions for privacy-preserving speech processing applications such as speaker verification, speaker identification, and speech recognition. The thesis introduces tools from cryptography and machine learning and current techniques for improving the efficiency and scalability of the presented solutions, as well as experiments with prototype implementations of the solutions for execution time and accuracy on standardized speech datasets. Using the framework proposed may make it possible for a surveillance agency to listen for a known terrorist, without being able to hear conversation from non-targeted, innocent civilians.