About me

As you already my know (or not) I am a Telecommunication engineer. I finished it in March 2008. Then I did a Master in Information technologies and communications in wireless networks, which I also finished in September 2009.

What am I working on now?

Currently I am granted by the Aragón Government for research staff and doing my PhD. I am part of Voice Group at University of Zaragoza (http://www.vivolab.es GTC, CPS, Universidad de Zaragoza)

Right now, I am working on speaker and segmentation and clustering, mainly evaluating how the bayesian information criteria works with different voice parameters.
Maybe it sounds like chinese, but basically is detecting whether a person or another is speaking at a given time.

If you are curious, my master's thesis is this one (Sorry, in spanish): memoria.pdf (Study of parametrization of voice signal using AM-FM model for speaker segmentation and clustering.

Abstract:

Resonant voice signals have amplitude and frequency modulation information. By means of using energy operators we achieve an efficient demodulation. In this thesis this method will be used for obtaining a set of parameters, which have been previously used for phoneme recognition. What we are going to do is to apply them to the segmentation task for splitting different audio sections, grouping segments with common characteristics and finally we will evaluate the amount of information needed for improving the speaker segmentation and clustering system.

Previous work:

My final project consisted of designing a digital crossover filter system for speakers and line-arrays, in collaboration with a company from Huesca. The system was getting good but finally the development got stuck, maybe because of the lack of i+d funding.

You may download this work also here (yep, spanish sorry): Final project, anexos.pdf (Design and analysis of digital crossover filters for speaker characterization)

The main objective is to implement a design, measuring and validation system for speakers and an automatic EQ. For this a design of a set of IIR and FIR filters has to be done, taking into account the signal to be processed is audio. Besides, an efficient implementation of filters has to be taken into account (real time).