You are welcome to browse through our demos and videos (alphabetical order). A broad and non-exhaustive list of the team’s research topics may be found on our homepage. Some of these demos are related to our research pages and are directly linked to recently submitted or accepted publications that can be found here. Please also refer to our complete list of publications.
Demos and Videos
Audio-visual multiple-speaker tracking
We exploit the complementarity of audio and visual information for tracking multiple persons and for assigning segments of speech to each person, over time. The tracker is based on a variational Bayesian formulation which yields a computationally tractable solution. Please visit our research page for more details. Acknowledgments: Work funded by the European Union under …
Audio-visual person tracking with NAO
This video summarizes some of the work carried out by the Perception team in 2017. We use a NAO robot manufactured by Softbank Robotics Europe. Unlike the standard (commercial) version, our NAO has a stereoscopic camera pair which allows to track persons in 3D and to implement visual servoing robustly and efficiently. The vides shows …
Audio-visual speaker diarization
Speaker diarization consists of assigning speech signals to speakers engaged in dialog. We proposed audio-visual spatiotemporal diarization model that tracks multiple persons and assigns acoustic signals to each person (please visit our research page for more details). Below are some of our results on AVDIAR dataset. The digit displayed on top of a person head …
Audio-visual tracking, speaker diarization and speech recognition
This video summarizes some of the work carried out by the Perception team in 2018. The video shows multiple person tracking, audio-source localization, audiovisual alignment, speaker diarization, as well as a complete pipeline, including the assignment of segments of speech to persons, and speech recognition. Acknowledgments: Work funded by the European Union under the ERC …
Eye gaze and visual focus of attention
This video uses examples from the LAEO dataset to illustrate our method for tracking gaze and visual focus of attention. Red arrows indicate the head orientations and green arrows indicate gaze directions. A circled face indicates the visual focus of attention of a person gazing towards that face. A dashed line between two faces indicates …
NAOLab
A Distributed Architecture for Interacting with NAO NAOLab is a middleware library for developing robotic applications in C, C++, Python and Matlab, using the humanoid robot NAO Software Download | Publications | People | Support | Acknowledgements NAOLab is a middleware for the development of robotic applications in C, C++, Python and Matlab, using the humanoid robot NAO …
