Reza Yazdani Aminabadi

PhD Researcher


Welcome to my website

I am a PhD researcher at Computer Architecture department of Universitat Politecnica de Catalunya. I study my PhD under the supervision of Jose-Maria Arnau and Antonio Gonzalez at ARCO (ARchitecture and COmpiler) group. My current research is focusing on designing energy-efficient and high-performance hardware accelerators for Automatic Speech Recongnition (ASR) systems.

I started my PhD at 2015. I specifically work on different stages of ASR pipeline which are mainly the acoustic likelihood computation (DNN) and speech graph search (Viterbi Search). For now, I have mostly studid the second phase, Viterbi Search, as it is the main bottleneck of these systems. So far, I have implemented an accelerator which outperforms all the previous designs and also consumes much less energy than others. Our accelerator is meant to remain the same or even perform better as the speech models gets more complicated and huge vocabulary is used.


News

  • Our new Idea of using on-the-fly composition algorithm for the ASR Systems has been accepted in the MICRO of this year. It will be presented on October 2017 in Boston, US.
  • Exploring DNN and RNN for the hybrid speech recognition systems
  • Increaing the efficiency of ASR systems through partitioning the speech recognition grpah (WFST). Reducing the size of speech data-set by an order of magnitude!!