BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260803T031216EDT-8166m8feuN@132.216.98.100
DTSTAMP:20260803T071216Z
DESCRIPTION:Acoustic\, perceptual\, and cognitive features for objective\ns
 peech quality and intelligibility prediction: Dr. Tiago Falk (INRS\n- EMT)
 . A CRBLM symposium\nAbstract:\nSubjective listening tests are commonly us
 ed by\ntelecommunications service providers to gauge network performance\;
 \nby speech language pathologists to characterize speech disorder\nseverit
 y or to document treatment outcomes\; or even by clinicians\nlooking to de
 tect diagnostic “markers” in e.g.\, autism spectrum\ndisorders (ASD). Subj
 ective listening tests\, however\, are costly\,\nlaborious\, and subject t
 o many intrinsic variables and biases due\nto e.g.\, familiarity with the 
 patients and their speech. Objective\nquality prediction\, on the other ha
 nd\, replaces the human listener\npanel by a computer algorithm\, thus pro
 vides an economical and\nreliable (repeatable) alternative.\n\nThis talk w
 ill describe our efforts at developing\nbiologically-inspired signal proce
 ssing algorithms that extract\nrelevant acoustic and perceptual features f
 or automated speech\nquality and intelligibility assessment. Three applica
 tions of the\ndeveloped algorithms will be presented. First\, I will prese
 nt\nfindings on an auditory-inspired quality model for natural speech\nwhi
 ch was recently shown to be particularly useful for cochlear\nimplant user
 s in reverberant environments. Second\, I will show that\nby characterizin
 g and integrating multiple dimensions (e.g.\,\nprosody\, temporal dynamics
 \, vocal source excitation)\, it is\npossible to predict the intelligibili
 ty level of a given dysarthric\nspeaker as well as predict the success of 
 such speaker in using an\noff-the-shelf speech recognizer for communicatio
 n purposes. Lastly\,\nI will present some pilot findings on the use of aut
 omated acoustic\nfeature analysis for the purpose of diagnosing ASD risk i
 n\nnon-verbal toddlers. Time permitting\, I will describe an ongoing\nproj
 ect which aims at developing“cognitive” features for objective\nspeech qua
 lity perception\, motivated by some recent findings in our\nbrain-computer
  interface research.\nAbout Dr Tiago Falk:\nTiago H. Falk received the BSc
  degree from the Federal University\nof Pernambuco\, Brazil\, in 2002\, an
 d the MSc and PhD degrees from\nQueen's University\, Canada\, in 2005 and 
 2008\, respectively\, all in\nelectrical engineering. In 2007\, he was a v
 isiting Research Fellow\nat the Sound and Image Processing Lab\, Royal Ins
 titute of\nTechnology (KTH)\, Sweden\, and in 2008 at the Quality and Usab
 ility\nLab\, Deutsche Telekom\, Germany. From 2009-2010 he was a\nPostdoct
 oral Fellow at Holland-Bloorview Kids Rehabilitation\nHospital\, affiliate
 d with the University of Toronto. Since Dec.\n2010\, he has been an Assist
 ant Professor at the Institut National\nde la Recherche Scientifique (INRS
 ‐EMT) in Montreal where he is\nDirector of the Multimedia/Multimodal Signa
 l Analysis and\nEnhancement (MuSAE) Laboratory. Dr. Falk is also an Adjunc
 t\nScientist at Holland-Bloorview and a Research Advisor for InteraXon\nTh
 ought Controlled Computing. His research interests include speech\nquality
  measurement and enhancement\, neural correlates of speech\nquality percep
 tion\; assistive technologies\; speech communication\ndisorders\; and huma
 n‐machine interfaces. He has published over 70\njournal manuscripts\, conf
 erence papers\, and book chapters in these\ntopics.\n\nMore about Dr. Falk
 : http://www.inrs.ca/english/tiago-h-falk\n
DTSTART:20120118T203000Z
DTEND:20120118T220000Z
LOCATION:CA\, 1160 Pine Ave. West\, 1160 Pine Ave. West\, Room 501\, Goodma
 n Cancer Centre\, Montreal\, QC
SUMMARY:Acoustic\, perceptual\, and cognitive features for objective speech
  quality and intelligibility prediction: Dr. Tiago Falk
URL:/channels/event/acoustic-perceptual-and-cognitive-
 features-objective-speech-quality-and-intelligibility-prediction-d-213302
END:VEVENT
END:VCALENDAR