Demo: video annotation tool, Cédric Fayet
Talk: Semantic analysis using frames, Gabriel Marzinotto
Talk : Still words, always words : from event’s date detection to song lyrics, Rémi Kessler
Talk: incremental learning of domain-specific models through dialog interaction, Vincent Letard
Best paper award at TALN 2017
The paper of Raheel Qader, Gwénolé Lecorvé, Damien Lolive, and Pascale Sébillot on speech disfluencies got rewarded by the Best Paper Award at the conference TALN 2017, held in Orléans (France) in June. The paper can be accessed here: https://hal.inria.fr/hal-01532031/file/disfluency-generation.pdf.
Annual team meeting 8-9 June
Programme Keynote speaker : Laurent Amsaleg (IRISA/CNRS, team LinkMedia, “multidimensional indexing and search by similarity: large scale, speech, quality”. Speakers : Pamela Carreno, Antoine Perquin, Raheel Qader, Aghilas Sini, Cédric Fayet, Rémy Kessler, Ahmad Alaa Eddine, Stefania Pecòre. Demo : Claude Simon, Damien Lolive.
Dana Kulic Seminar, Motion analysis and segmentation, April the 28th.
Dana’s work, as lead of the Adaptive Systems Laboratory at the University of Waterloo, CA, has focused on the analysis of human movement in a variety of applications, including human-robot interaction, rehabilitation and sports training, and interactive architectural and sculptural installations.
PhD defense : Marc Dupont: Glove-based gesture recognition for real-time outdoors robot control, 28th March, 2017.
Although gesture recognition has been studied for several decades, much research stays in the realm of indoors laboratory experiments. In this thesis, we address the problem of designing a truly usable, real- world gesture recognition system, focusing mainly on the real-time control of an outdoors robot for use by military…
PhD defense : Raheel Qader: Pronunciation and disfluency modeling for expressive speech synthesis, 31st March, 2017.
In this thesis, we address the problem of expressivity in Text-To-Speech (TTS) by incorporating two phenomena with a high impact on speech: pronunciation variants and speech disfluencies. In the first part of this thesis, we present a pronunciation variant generation method which works by adapting standard, i.e., dictionary-based, pronunciations to…
