[Corpora-List] Errors by Humans and Machines in multimedia, multimodal and multilingual data processing – ERRARE 2015 - online program and webcast

Verginica Mititelu vergi at racai.ro
Mon Sep 7 13:50:30 CEST 2015

Errors by Humans and Machines in multimedia, multimodal and multilingual data processing – ERRARE 2015

11-13 September 2015, Sinaia (Romania)

The Research Institute for Artificial Intelligence “Mihai Draganescu” (ICIA) of the Romanian Academy in collaboration with IMMI-CNRS, LIMSI-CNRS, the LAbEx EFL and A.I Cuza University of Iasi, organizes the second edition of the “Errare” workshop, in September 11-13, 2015, as a satellite event of Interspeech 2015 (http://errare2015.racai.ro). "ERRARE 2015" will be webcasted (http://er-tube.mediaec.uaic.ro/).

PROGRAM (also available here: http://errare2015.racai.ro/conference.html#program).

Saturday, 12th of September 9:00-9:15 Welcome speech by Acad. Florian Filip, Romanian Academy, Chair of the "Information Science and Technology" Section

Presentation of the workshop, Ioana Vasilescu

9:15 Session #1 Variation and speech: analysis and modeling of errors in speech-to-text applications

Session chair: Gilles Adda

9:15-10:15 Keynote C. Burileanu: “The Romanian ASR System developed at “Speech & Dialogue” (“SpeeD”) Research Laboratory – University POLITEHNICA of Bucharest. Under-resourced languages challenges for an ASR system”

10:15-10:30 Break

10:30-11:00 D. Telaar, J. Weiner, T. Schulz: “Error signatures to identify Errors in ASR in an unsupervised fashion” 11:00-11:30 M. A. Ben Jannet, M. Adda-Decker, O. Galibert, S. Rosset: “How to evaluate ASR errors impact on NER?” 11:30-12:00 H. Cucu, A. Buzo, C. Burileanu: “ASR errors in transcribing informal pronunciations of Romanian numbers” 12:00-12:30 S. Ghannay, N. Camelin, Y. Estève: “Which ASR errors are hard to detect?”

12:30-13:45 Lunch 13:45 Session #2 From Variation to Errors in Speech Production and Perception

Session chair: Martine Adda-Decker

13:45-14:45 Keynote R. Hartsuiker: “Conflict monitoring in speech production” 14:45-15:15 S. Marin, M. Pouplier: “Spontaneously occurring speech errors in German: BAS corpora analysis” 15:15-15:45 C. Dutrey, M. Adda-Decker, F. Santiago: “Large-scale prosodic-acoustic analysis of ASR errors in French”

15:45-16:00 Break

16:00-17:00 Keynote M. Zock: “« Errare humanum est ». Refusing to « appreciate » this fact could be a big mistake!” 17:00-17:30 L. ten Bosch, G. Giezenaar, L. Boves, M. Ernestus: “Modeling language-learners’ errors in understanding casual speech” 17:30-18:00 I. Chitoran, J. Hualde, O. Niculescu: “Gestural undershoot and gestural intrusion - from perception errors to historical sound change”

Sunday, 13th of September 11:00 Session #3 Errors and dialogue: issues in Human/Human and Human/Machine Interaction

Session chair: Joseph Mariani

11:00-12:00 Keynote F. Lefèvre: “Uncertainty management in vocal interaction interfaces, seamlessly modeling and handling errors in interactive processes” 12:00-12:30 R. Higashinaka, K. Funakoshi, M. Mizukami, H. Tsukahara, Y. Kobayashi, M. Araki: “Analyzing dialogue breakdowns in chat-oriented dialogue systems”

12:30-13:45 Lunch

13:45-14:15 L. Devillers, S. Rosset : “The communication accommodation of the machine to deal with errors” 14:15-14:45 A. I. Niculescu, R. Banchs: “Strategies to cope with errors in human-machine speech interactions: using chatbots as back-off mechanism for task-oriented dialogues”

14:45-15:00 Break

15:00 Session #4 Errors from multiple sources (syntax, discourse, multimedia...)

Session chair: Dan Tufis

15:00-16:00 Keynote M. Turchi: “Working with human post-edited sentences: old challenges, new solutions.” 16:00-16:30 M. Kai : “Truncated Speech and Its Repair” 16:30-17:00 F. Santiago, M. Adda-Decker, C. Dutrey: “Towards a

Typology of ASR Errors via Syntax-Prosody Mapping”

17:00-17:15 Break

17:15-17:45 V. Barbu Mititelu, E. Irimia: “Types of errors in the automatic syntactic parsing of Romanian” 17:45-18:15 D. Charlet, J. Poignant, H. Bredin, C. Fredouille, S. Meignier : “What Makes a Speaker Recognizable in TV Broadcast? Going Beyond Speaker Identification Error Rate”

18:15-18:45 Final discussion

More information about the Corpora mailing list