Interspeech--the world’s largest and most comprehensive conference on the science and technology of spoken language processing -- is coming to India for the first time ever. This year will witness the 19th edition of Interspeech, which will be held during September 2-6, 2018 at the Hyderabad International Convention Centre (HICC), Hyderabad, India, with an expected international participation of over 1500 delegates.
The conference, being organised by the International Speech Communication Association (ISCA), will bring together researchers from across the globe to discuss and showcase the state-of-the-art technologies in speech recognition and related areas. “It is a conference about speech, acoustics, audio, language, machine learning, and artificial intelligence. It goes into multiple domains including technology, science and applications of these areas, especially in the context of mobile communication,” says Dr. Prasanta Kumar Ghosh from the Indian Institute of Science (IISc), who is also a Technical Program Chair of the conference.
Interspeech 2018 will feature plenary talks by internationally renowned experts, tutorials, show & tell sessions and exhibits. The general theme for this edition of the conference is ‘Speech Research for Emerging Markets in Multilingual Societies.’ “Starting from the basic understanding of speech, like how humans speak and the anatomy involved, the conference delves into the engineering aspects (speech processing and recognition), language aspects (translation software), and health aspects (speech disorders), among many others,” says Dr. Ghosh.
You might have heard people shout "Hey Alexa", "Okay Google", and "Hey Siri" into their 'smart' devices and trying to get these virtual assistants to do something. Amazon’s Alexa, Google's Assistant and Apple's Siri are voice-based virtual assistants that can respond to your voice commands. They can make calls, send text messages, provide updates on the weather, fix appointments and much more! But how do they do that? It is through speech recognition and understanding software that recognises a set of preset ‘commands.’
A common challenge in developing such speech recognition software is understanding the different accents people use. In a multilingual country like India, the vast diversity of languages presents a bigger challenge. The conference tries to address this and also showcases the latest advances in spoken language and communication research.
“The theme for this year's conference has been chosen keeping the Indian context in mind. In the past, there has been little participation from Indian students and industry. But since it will be happening in Hyderabad this year, which is easily accessible, we are expecting a larger participation from across the country,” remarks Prof. Chandra Sekhar Seelamantula, an Associate Professor at IISc, who is also the Chair of the Publicity Committee of the conference.
The conference is supported by the International Speech Communication Association and is organized by “The Who’s Who of Speech Research in India” from IISc, IIT Madras, IIT Bombay, IIT Guwahati, IIT Hyderabad, IIT Kharagpur, IIT Mandi, IIT Kanpur, IIIT Hyderabad, IIIT Bangalore, NITs, AIISH Mysore, as well as from top corporate companies such as Microsoft, Amazon, Samsung, IBM Research, LinkedIn, and leading institutes from across the world, working cohesively under the leadership of Prof. B. Yegnanarayana, the father figure of speech research in India.
The conference participants include students, academicians, and industry representatives. The Amazon Alexa group is a founding member and sponsor of the event. Other sponsors and supporters include Uniphore, Microsoft, Baidu, Samsung, Tencent, Qualcomm, Google, Nvidia, GoVivace, University of Washington, Vignan University, BVRIT and the Indian Institute of Science. Many Government and R&D organizations as well as industries in India are expected to play a major role in terms of sponsorship and participation in this conference. There are many opportunities for startups as well as established industries and educational institutions to showcase their activities through exhibition stalls.
Since this is the first time Interspeech is coming to India, the expectations are high. “The speech processing industry was predominantly in Europe and US. In early 2000, the conference was held in China, after which China has produced many research papers in this area. We hope a similar story will unfold in India as well,” says Dr. Sriram Ganapathy from IISc, who is on the Sponsorship Committee of the conference.