Center for Forensic Phonetics and Acoustics (CFPA)

Mission

The Center for Forensic Phonetics and Acoustics brings together an interdisciplinary array of research from diverse fields of phonetics, psychology, medicine, engineering, physics in academia as well as forensic industries to address current questions in voice recognition relevant for forensic investigation such as: 

  • How individual are voices? 
  • How do voices vary between individuals?
  • How do humans recognise voices?
  • What control do speakers have over voice identity? 
  • How can voices be recognised automatically? 
  • How secure are automatic voice detection systems?

What we do:

Research: Forensic services: Education:
  • Speech timing
  • Automatic recognition
  • Shouted voices
  • Voice disguise
  • Voice line-ups
  • Speaker profiling
  • Speaker comparison
  • Transcription
  • Questioned utterances
  • Authentication
  • Audio enhancement

(for prosecution & defence)

  • Students/PhD candidates
  • Police
  • Prosecutors
  • Defence lawyers

Executive Board

Board_Pic

Members

Ph.D. in General Linguistics; research in forensic phonetics, in particular exploring the speaker-specific acoustic characteristics in Persian and other languages and dialects of Iran. 

  • Honglin Cao (Prof. Dr.; professorial fellow; webpage)

Ph.D. in Phonetics; forensic expert with a specialisation on the face-voice relationship.

  • Volker Dellwo (Prof. Dr.; webpage)

Ph.D. in Phonetics; research in speech timing, and voice dynamics with strong links to forensic applications.

  • Kostis Dimos (Ph.D. candidate; webpage)

Ph.D candidate in Phonetics; research on speech rhythm perception and production, emotional prosody, the phonetics-phonology interface and shouted voices. 

  • Peter French (Prof. Dr.; professorial fellow; webpage

Chairman of JPFrench Associates, one of the leading experts in forensic voice evidence worldwide.

  • Thayabaran Kathiresan (Ph.D Candidate, webpage)

Ph.D. candidate in phonetics and speech sciences; research in indexical variability in speech register changes in automatic speaker recognition systems.

Ph.D. in Linguistics; research in models of speech rhythm, differences between speakers, and articulatory measurements.

Ph.D. in Linguistics; research in Swiss German dialectology with forensic phonetic applications.

  • Omnia Ibrahim (Ph.D. candidate; webpage)

Ph.D. candidate in Linguistics/Phonetics; research human-computer voice interaction and developmental aspects of accommodation

  • Ji-Eun Kim (Ph.D. candidate; webpage)

Ph.D. in Korean Linguistics; research in the relationship between culture and voice attractiveness and dialect variation of intonation.

PhD in Computer Science (2003), research interests in audio and speech processing (coding, speech and speaker recognition), conversation analysis and machine learning. Many of the designed applications are developed in collaboration with security/government (LEA) bodies in Switzerland, or at EU level.

  • Elisa Pellegrino (Dr. M.A.; webpage)

Ph.D. in Linguistics; research in phonetic convergence in speech communication,  speaker-specific factors (language background, age and health condition) affecting speech rhythm variability.

Ph.D. in Linguistics;  research in acquisition of prosody in the second language, speech rate perception and production in L1 and L2, and computer-assisted pronunciation training.

  • Radek Skarnitzl (Prof. Dr., webpage)

Ph.D. in Phonetics; research interests in forensic phonetics and socio-psychological aspects of speech perception.

Ph.D. candidate in Phonetics; research in between-speaker variability in tone-languages. 

  • Urban Zihlmann

Ph.D. candidate in German linguistics; research on phonetics and Swiss German dialectology applied to forensic casework.

Selected Bachelor & Master Research Projects

  • Dario Brander (M.A.; 2013) The use of hesitation for forensic phonetic speaker comparison
  • Lukas Fischer (M.A.; ongoing) Voice variability with age
  • Andrea Fröhlich (M.A.; 2018) Evaluation des forensischen Sprechererkennungssystems 'iVocalise' 
  • Sarah Lim (M.A.; ongoing) Speaker recognition in deceptive speech
  • Elena Mayorova (M.A.; ongoing) Temporal information in speaker recognition
  • Linus Manser (B.A.; 2018) Automatic age recognition
  • Sibylle Sutter (M.A.; 2012) The influence of visual speech on auditory speaker identification ability

Selected Funded Research Projects

  • Dellwo, V. (2014-2016) VoiceTime. [Gebert-Rüf Foundation; GRS-027/13]; 340'000 CHF.
  • Dellwo, V. (2011-15) The functions of speech rhythm in segregating simultaneous speakers [Stiftung für wissenschaftliche Forschung der U. Zürich]; 14'000 CHF.
  • Dellwo, V./Frühholz, S. (2018-2020) Voice Theft: Chances and risks of digital voice technology [Swiss National Science Foundation; 10DL15_183152]; 250’000 CHF.
  • Dellwo, V./Kerdpol, K. (2017-2018) Speaker recognition in tone languages. [State Secretary for Education, Research and Innovation (SERI); seed money 2013-16]; 10'000 CHF.
  • Dellwo, V./ Schmid, S. (2012-2015) Speaker identification based on speech temporal information: A forensic phonetic study of Swiss German [Swiss National Science Found.; 100015_135287]; 432'456 CHF.
  • Dimos, K./ He, L./Dellwo, V. (2015-2016) An investigation of the rhythmic acoustic differences between normal and shouted voices [International Association for Forensic Phonetics and Acoustics]; 1'800 GBP.
  • He, L./Pellegrino, E./Dellwo, V. (2017-2018) Forensic phonetic analysis of electronically disguised voice [International Association for Forensic Phonetics and Acoustics]; 1’800 GBP.
  • Schwab, S./Dellwo, V. (2017-2018) Effects of nicotine on voice quality [International Association for Forensic Phonetics and Acoustics]; 1’800 GBP.

Selected Recent Publications

  • Asadi, Homa ; Nourbakhsh, Mandana ; He, Lei ; Pellegrino, Elisa ; Dellwo, Volker (2018).  Between-speaker rhythmic variability is not dependent on language rhythm, as evidence from Persian reveals.International Journal of Speech, Language and the Law, 25(2):151-174. ZORA
  • Braun, A., Llamas, C., Watt, D., French, J. P., & Robertson, D. (2018). Sub-regional ‘other-accent’ effects on lay listeners’ speaker identification abilities: a voice line-up study with speakers and listeners from the North East of England. International Journal of Speech, Language and the Law25(2), 231-255. https://doi.org/10.1558/ijsll.37340
  • Dellwo, Volker ; French, Peter ; He, Lei (2018). Voice Biometrics for Forensic Speaker Recognition Applications. In: Frühholz, Sascha ; Belin, Pascal . The Oxford Handbook of Voice Perception. Oxford: Oxford University Press, 777-798. ZORA
  • Foulkes, P., French, P., & Wilson, K. (2019). LADO as forensic speaker profiling. In P. Patrick, M. Schmid, & K. Zwaan (Eds.), Language Analysis for the Determination of Origin SPRINGER.
  • Kolly, Marie-José ; Boula de Mareüil, Philippe ; Leemann, Adrian ; Dellwo, Volker (2017).  Listeners use temporal information to identify French- and English-accented speech. Speech Communication, 86:121-134. ZORA
  • He, Lei ; Dellwo, Volker (2017).  Between-speaker variability in temporal organizations of intensity contours. Journal of the Acoustical Society of America, 141(5):EL488-EL494. ZORA
  • He, Lei ; Dellwo, Volker (2016). The role of syllable intensity in between- speaker rhythmic variability. International Journal of Speech, Language and the Law, 23(2):243-273. ZORA