PHASE III

Presenter Name	Ashray Malhotra, Nisheeth Lahoti
Cluster	EE1
Email-id:
Topic:	SoundRex - Next generation of Sound Technology for concerts.
Venue:	GG 301
Session Chair:	Nitya Tiwari
Date:	03/08/2016 at 4 PM
Pre-Requisite:	None
Abstract:	At SoundRex, we are building the next generation of sound technology. The way sound technology has evolved historically is that we started with a mono speaker, which is just a single sound source. Then we went to stereo speakers, which are 2 speakers. The commercial state of art is 5.1, where you have 5 speakers. But in concerts, which are supposed to be the best musical experience of your lives, we still have mono speakers and at best stereo speakers. What we have built is radically different. What we propose is that everyone who enters the concerts gets a wearable wrist band speaker. Which means that if there are 10k people in a concert, you are hearing sound from not just 2 speakers in front of you but from thousands of speakers distributed all across the concert. This gives the listeners an unprecedented feeling of being immersed in sound since the sound is coming from all directions and not just the front. Also since we have control over thousands of speakers and not just 2, this allows us to create completely new sound effects which are not even possible with just 2 or 5 speakers. In this talk, we will give an overview of the technology challenges that we have faced in building this technology ground up. We will also cover some general aspects with building a technology (especially hardware) startup in India. Also, we will highlight what it takes to build a real product for the masses in contrast to a lab prototype. Do join us, we hope to have an interactive session.
Presentation Slides:

Presenter Name	Nitin Bhatia, Ph.D, Fibre Optics Lab
Cluster	EE1
Email-id:
Topic:	There's always Light at the end of the Tunnel.
Venue:	GG 301
Session Chair:	Thomas Joseph
Date:	05/10/2016 at 4.30 PM
Pre-Requisite:	None
Abstract:	What comes out of an optical fiber is a Light beam. What else one can expect? It seems that the cylindrical fiber geometry does a nice trick to provide us with exotic beam patterns with special optical properties. In this talk, we will have a glimpse of some of the exciting properties of these light beams, termed as Bessel and Vortex beams, and their wide range of application areas including optical trapping, fiber lasers and optical communication systems.
Presentation Slides:

Presenter Name	Aditya MVS, Ph.D
Cluster	EE1
Email-id:
Topic:	Price Competition in Spectrum Markets: How accurate is continuous price approximation ?
Venue:	GG 302
Session Chair:	Amit More
Date:	07/10/2016 at 4 PM
Pre-Requisite:	None
Abstract:	Primary users (like Reliance, Airtel) of the spectrum compete solely to sell the under-utilized spectrum to secondary users (Iike BPL). Literature assumes that a primary user can choose any real number as a selling price of the spectrum, for the ease of mathematical analysis. In practice, however, prices can only be in multiples of say a rupee/paise/dollar/cent. The fundamental question of How accurate are the results, when a discrete problem (actual price) is approximated by a continuous problem (price as a real number) in spectrum markets? is investigated.
Presentation Slides:

Presenter Name	Avik Hati Ph.D, Vision and Image Processing Lab
Cluster	EE1
Email-id:
Topic:	Saliency: Where do you look in the image?
Venue:	GG 303
Session Chair:	Amit More
Date:	21/10/2016 at 4 PM
Pre-Requisite:	None
Abstract:	Saliency is a measure of importance of objects in an image or events in a video scene. The image or scene elements that capture our attention are salient. They are distinctive in color or spatial features from rest of the image or temporal segments. Detection of salient regions in images or videos aids in many image and vision applications e.g., object segmentation, video summarization, content based image compression, image and video quality assessment etc., as only the important parts may be processed. This leads to reduction in complexity as these applications work with large image databases and long video sequences. This talk will be limited to image saliency detection.
Presentation Slides:	Click Here.