While there was a brief decrease in the interest of voice applications over the past few years, today with the advent of eye-garnering products like Alexa, Siri and Cortana, there is no doubt that voice-enabled technologies are experiencing a resurgence. With a long history of research work done in this area, it has been implemented in various biometric identification and forensics applications in the past.
These usages can be, in general, interpreted as speaker identification or speaker verification tasks. It analyzes the acoustic features of speech that have been found to differ among individuals. Speaker recognition is a task of identifying people from their voices. Darshan Kedari (181060021), Mr.Anveshak Rathore (181060012), Mr.Vedant Mankar (181060040), Mr.Sourav Sawhney (181060001), Mr.Dixit Mendon (181060043), is found to be satisfactory and is approved for partial fulfillment of the Degree B.Tech Electronics The Project Titled " A Comparative Study of GMM and SincNet models for Speaker Identification" by Mr. Rathore Head of Electrical Engineering, Dept This is to certify that Darshan Kedari (181060021) Anveshak Rathore (181060012), Vedant Mankar (181060040), Sourav Sawhney (181060001), Dixit Mendon (181060043),students of B.Tech Electronics, Veermata Jijabai Technological Institute, Mumbai has successfully completed the Project titled " A Comparative Study of GMM and SincNet models for Speaker Identification" and submitted a report on the same to my satisfaction. Rizwan Ansari for giving unsurpassed suggestions towards successful completion of this new project. We are very much thankful to all other teaching and non-Teaching Staff members of the Electrical Department, especially our mentor Dr. We have to express our gratitude to, Head of Department, Electrical Department, for his invaluable motivation and encouragement, which helped us in the whole process of this project. Also we are vastly obliged to him as our project guide, for his breed and valuable support to make our project a successful one. Rathod for giving valuable guidance, steady support and encouragement to inclusive our project lucratively. We are very glad to expose our sincere and lovable memorial thanks to our management for having on hand the facilities for the triumphant completion of the project. The Report reflects all the work done during the entire time of dissertation. Veermata Jijabai Technological Institute (VJTI) and have not been submitted anywhere else. Rathod at the Department of Electrical Engineering. We, hereby declare that the Project titled " A Comparative Study of GMM and SincNet models for Speaker Identification" being submitted by us towards the partial fulfillment of the Degree of B.Tech Electronics is a project work carried by us under the guidance of Dr.
23th May 2008 : Version 4.20 - Added: Mail library - Added: Ftp library - Added: RegularExpression library - Added: Http library - Added: SerialPort library - Added: FLAC library - Added: SQLite library - Added: DirectX 9 subsystem - Added: Sound Streaming for OGG and FLAC - Added: #PB_Sound_MultiChannel support for the sound library - Added: CreateImageMenu(), CreatePopupImageMenu(), SHA1FingerPrint(), SHA1FileFingerPrint() - Added: CRC32FileFingerprint(), ExamineMD5FingerPrint(), ExamineSHA1FingerPrint(), - Added: NextFingerPrint(), FinishFingerPrint(), IsFingerPrint() - Added: Hexadecimal and binary support for Val() and ValQ() - Added: #PB_Compiler_Procedure compiler constant - Added: #PB_Round_Nearest support for Round() - Added: 'XOr' drawing support for Linux/SDL and MacOS X - Added: New "Profiler" tool for the Debugger menu to monitor how often lines are called in the code - Added: Debugger can now evaluate expressions (see Debug output) - Added: The Memory Viewer can accept any expression as the start/end address - Added: Adding a "+" at the "end address" field of the Memory viewer makes the end address relative - Added: Placing the mouse over a Variable will display its value while the program runs.Project Report titled A Comparative Study of GMM and SincNet models for Speaker Identification