Manolis Perakakis world

News, diary, journal, whatever

Prime time for Distributed Speech Recognition? February 23, 2009

While an undergraduate student a few years ago I worked on Distributed Speech Recognition (DSR). The main purpose of DSR is to compress the acoustic features used by a speech recognizer, over a data (instead of voice) network, thus saving bandwidth (cost effective) and allowing the use of full speech recognition in mobile terminals. As it compresses acoustic features for speech recognition (not speech signal transmission/reproduction) purposes it can achieve very low bit rates. You can think of it as analogous of what mp3 is for music transmission and storage.

Depicted next is a simple overview of a DSR architecture (model 2). Note that the mobile terminals depicted are Symbian’s reference devices corresponding to smartphone, handheld and PDA respectively (Ooops too old images – it should be back in 2001; should upgrade to something like iPhone or Android …)

My work with Prof. V.Digalakis concluded that one can successfully take advantage of DSR with only a 2 kbps coding, which is an extremely low data rate. After that i ported the DSR engine to a Zaurus Linux PDA and made it work in real-time (a 16MB, 200 MHz StrongArm processor).

Although my recent work focus is now on Multi-modal (speech) interfaces I still keep an eye on DSR. It seems that with the emergence of powerful mobile terminals and the announcement of speech recognition support for Android and iPhone by Google, DSR might become soon a hot topic!

P.S. I just found out my DSR page is ranked 3rd by Google after W3C and ETSI. Holy moly!

Coolness factor: ?

Advertisements
 

Academic publishing February 21, 2009

Filed under: academic research — perak @ 4:49 am
Tags:

Luis Von (CMU professor & inventor of CAPTHA) points out in his post Academic publishing 2.0 how the academic research world has been turned to a paper-generation industry:

As an academic community, it sometimes feels that the final goal of doing research is publishing papers. The goal of doing research should be, well, doing research. I understand that communicating the results of our work is important, but surely there is a better method than one that was invented before computers were around.

Given the number of people working in computer science and the fact that publishing papers is considered the goal of our work, there is an insane number of papers written every year, the vast majority of which contribute very little (or not at all) to our collective knowledge. This is basically spam.

Can a combination of a wiki, karma, and a voting method like reddit or digg substitute the current system of academic publication?

Well, this is basically a chicken and egg loop problem. Many people doing research know they need to produce a large number of research papers each year to stay competitive. Quantity over quality. Although there are some efforts to quantify the quality of a research paper such as the impact factor, eigenfactor, or even using Pagerank the research public community needs to reinvent itself!

P.S. Darn, i need to finish some papers soon!

 

Free access to scientific knowledge

Filed under: academic research — perak @ 4:18 am
Tags:

I just got into PLOS the
public library to science site. Until now the focus of the library is on medicine, biology and genetics. It is so nice to see such efforts nowdays. Welcome to science 2.0!

I am sure there other other similar efforts out there. MIT, Berkeley and other top US universities publish their course lectures online. Sites such as academic earth and videolectures have a ton of really amazing video lectures.

As a electronic & computer engineer I never got it, that organizations like IEEE and ACM which should pioneer the access to technical advancements are so strict to sharing knowledge. I am sure they can come up with a smarter way of earning their revenues than just locking access to their papers. If they don’t any time soon it is likely they soon become obsolete.