IETE-SF organized a technical talk on speech and audio processing which was held on 20th February, 2018. Dr. Milind Shah (HOD, EXTC, Fr. Agnel College of Engineering, Vashi) was the speaker for the talk. Before proceeding with his session, he gave a brief introduction about himself and informed the audience about the topics which he would be covering during the session.


Dr. Shah kickstarted the technical talk by explaining about the speech chain. The biological process of speech creation in humans and the process of how human beings listen to sounds were discussed with the help of diagrams and loop chains. After enlightening the audience about the speech chain, he moved on to talk about visual and tactile feedback systems. With the help of equations and block diagrams, the concepts of frame based short time processing, generic short time processing and modified autocorrelation-based pitch estimation were thoroughly clarified by Dr. Shah. Topics like source-filter theory, techniques for vocal shaped tract and Wakita’s inverse filtering method were also covered during the session. In addition to theoretical knowledge, certain Matlab commands for pitch and shape estimation were made aware to the audience. Later, a brief introduction on spectrograms and their usage regarding sentence analysis was taught.

This technical talk was truly engaging and knowledge worthy. Dr. Shah’s session encouraged the third-year students among the audience to take up speech processing as their elective subject for the next semester whereas the final year students among the audience who have already opted for this subject as an elective gained additional knowledge apart from their regular curriculum awareness.