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: ?

My 15 minutes of fame! March 11, 2008

Our work in Telecommunications Lab, at Technical University Crete (TUC) was featured in “Orizontes” documentary series of Kydon TV channel. We demonstrated some of our demos :

• My work on multimodal interfaces (part of my PhD), including a travel. reservation multimodal (GUI + speech) application running on a Zaurus Linux PDA
• The automatic video summarizer system (part of MUSCLE NOE european research project showcases).
• An audio-visual (AV) recognition system (also part of MUSCLE NOE european research project showcases).
• The multi-mic robust speech recognition demo (part of Hiwire european research project showcases).

We could not showcase the augmented-reality demo, we developed in cooperation with VTT (speech recognition integration), since we currently miss the appropriate hardware, hope we get it soon.

Some of these demos will go public, either by posting videos on YouTube or by releasing the source as open source in Sourceforge/Google code.

More on this as well as a more detailed description of the demos in following future posts!

Stay tuned!

Aibo, Lego mindstorms, Wii remote (wiimote), iPhone & Google’s Android! March 11, 2008

What all these have in common? They will be my playground for a while …

I will have the chance to play with all of them during this samester!

As far as aibo and mindstorms are concerned, i will use them for the two robotics related courses i have enrolled in. Some possible projects I am thinking of :

• Distributed speech recognition (DSR) : enchance the limited speech recognition capabilities of the aibo by exploiting the wireless link and a  speech recognition server.
• Distributed image processing : enchance aibo’s limited machine vision capabilities by exploiting the wireless link and a machine vision server (similarly to DSR)
• robot localization using multiple input modalities : machine vision + audio
• enchanced gesture based interface or multimodal (speech + gesture interfaces)

Wiimote hacks for enchanced HCI, similar to these demos from CMU.

iPhone will be used,  to augment my speech & GUI multimodal interface prototype already  running on the Zaurus PDA, with the gesture modality.

Finally, i can’t resist from playing with Google’s  new Android platform,  for porting  various apps  I have in mind.

Whoa, my hacker alter ego will be definetely be back for good!!!