Giampiero Salvi:: Research
PhD Students
- Position available for PhD student in Machine Learning applied to Speech Technology. For more information click here
- Kalin Stefanov (co-supervised with Jonas Beskow and Hedvig Kjellström)
- Niklas Vanhainen
- Giovanni Saponaro (co-supervised with Alexandre Bernardino, Vislab, Institute for Systems and Robotics, Instituto Superior Técnico, Lisbon, Portugal)
- Alessandro Pieropan (co-supervised with Hedvig Kjelltröm, RPL, KTH, defended in 2015)
- Daniel Neiberg (co-supervised with Joakim Gustafson, defended in 2012)
- Chris Koniaris (co-supervised with Olov Engwall, defended in 2012)
- Gopal Ananthakrishnan (co-supervised with Olov Engwall, defended in 2011)
Audio-visual object-action recognition (KTH, 2013-present)
In this project we are studying how vision and hearing can be integrated to improve the ability of a robot to interpret human activities. We will show that in many cases the combination of vision and hearing will make it possible to disambiguate human actions.
Joint work with Alessandro Pieropan and Hedvig Kjellström
Free Resources for Automatic Speech Recognition (KTH, 1998-present)
I am a free software user, developer and enthusiast. I have long been working towards creating free resources for speech technology. Some of the software I contributed to develop is available here: http://www.speech.kth.se/software/ Free resources for Automatic Speech Recognition are available here: http://www.speech.kth.se/asr/
Joint work with Niklas Vanhainen
Biologically Inspired Automatic Speech Understanding (KTH, 2010-present)
The aim of the project is develop more flexible speech interfaces by drawing inspiration on how humans learn to interact by speech and language.
Joint work with Niklas Vanhainen and Gopal Ananthakrishnan
BioASU related Publications
Contact (IST, Portugal, 2005-2008)
As infants, did we learn to perceive and produce gestures for manipulation and speech independently, or are these two learning processes linked? Should robots? http://eris.liralab.it/contact/
SynFace (KTH, 2001-2004)
SynFace is a technology that allows an animated talking face to be controlled by the incoming speech signal in real time.
In this project I developed the low latency phoneme recogniser that permits to extract the information necessary to drive the face movements from the incoming speech in realy time.
This is the project home page, and this is the home page of the spin-off company SynFace AB
Synface related Publications