What about a system able not only to recognize who is speaking during a meeting but to understand the conversation and propose relevant documents like past reports, web pages as well as summaries and analyses of past meetings ? Artificial Intelligence and Speech Recognition specialists at Idiap are developing that smart automated assistant.
On January 25th, 2011, Dinesh Jayagopi successfully defended his PhD thesis entitled "Computational modeling of face-to-face social interaction using nonverbal behavioral cues".
He will work on neuron classification in videos, in a joint project with EPFL, University of Geneva, and University of Basel. More precisely he is going to study how to characterize neuron development dynamics, and their relation to gene expression.
On Dec 8th 2010, Sarah Favre successfully defended her PhD thesis entitled "Social Network Analysis for Automatic Role Recognition".
Every year, IMD searches for the most promising startups it can find to work with the MBA students. Over the past 13 years over 180 early-stage ventures have profited from this collaboration. Today, IMD representatives (Jim Pulcrano and Benoit Leleux) announce the twenty startups that have been selected to work with our MBA and EMBA students in 2011. Among them, Koemei (Idiap Spin-off) will participate to the venture.
On Dec 6th 2010, Deepu Vijayasenan successfully defended his PhD thesis entitled "An Information Theoretic Approach to Speaker Diarization of Meeting Recordings".
The 2010 Idiap awards have been granted to:
La journée "Osez tous les métiers" est organisée chaque année par le Secrétariat Valaisan à l'égalité et à la famille et a pour objectif d’aider filles et garçons à choisir leur métier au-delà de préjugés liés au genre.
For this competition we would like to measure the performance of algorithms that act as a counter-measure to spoofing.
At Idiap, Carl will be working on head pose estimation and visual focus of attention for the TA2 project.
Abdurrahman Eray Baran has joined Idiap as a PhD student under the supervision of Volkan Cevher, assistant tenure track professor
Sreeharsha has joined Idiap as a PhD student under the supervision of Dr. Fabio Valente.
Prof. Abutalebi will work with Prof. Bourlard in the framework of "Array-based Speech Signal Processing".
In November 15, Dr. Ronan Collobert will join the Idiap staff as permanent researcher.
Euresearch the Swiss guide to European research and support for European business and innovation highlights in its recent Success Stories Newsletter the EU FP7 project MOBIO (Mobile Biometry).
WFSTs represent a means of pre-compiling the grammars used by ASR decoders. Such pre-compiled grammars tend to lead to both quicker decoding, because duplication of weights and paths is minimised, and simpler decoder implementations, because the composition process is offline. Pre-compiled grammars are quite large, sometimes limited only by the size of the memory on the host CPU.
Data-flow is a signal processing paradigm where modular components can be arranged in directed graphs to perform complex tasks. Data-flow lends itself to block calculations that take advantage of the hierarchical caches on modern CPUs. It also naturally modularises flow-based algorithms such that the components are easily re-used.
Context modelling for object detection.