Bertrand Higy
– PhD Candidate –
Download my CV (pdf)

News

April-July 2018: Visiting the Centre for Speech Technology Research at University of Edinburgh for a 4 months period, to work on sequence-to-sequence speech recognition.

January 2018: Our paper Speech recognition for the iCub plateform has been accepted for a special issue of Frontiers in Robotics an AI.

30th Aug. - 8th Sept. 2017: Attending Barcelona Cognition, Brain and Technology summer school.

17th - 21st of July 2017: Attending Deep Learning summer school in Bilbao.

About me

Short biography

I achieved bachelor's degrees in Computer Science (2003-2006, first in my class) and Psychology (2005-2008) from Université Louis Pasteur, Strasbourg (France), followed by a master degree in Cognitive Sciences (2008-2010, first in my class) from Université Lumières, Lyon (France). From 2010 to 2015, I worked as a software engineer for a consulting company in Strasbourg. I am now part of the iCub Facility at the Italian Institute of Technology, where after 6 months as a fellow, working on haptic object recognition, I started a PhD.

Research topic

My current work focuses on deep learning for speech recognition, more precisely command recognition in the context of humanoid robotics. We want the robot iCub to recognize vocal commands addressed to him robustly, despite working in distant and noisy conditions. The low-resource condition in which we work make it harder to learn performant models for the specific context in which the robot operates.

Topics of interest

In the long run, my interest revolve around autonomous and developmental learning. I am particularly interested in unsupervised learning and computational modeling of language acquisition by children.

Projects

iCubRec

Command Recognition for the iCub plateform

Haptic Object Recognition

How to combine haptic modalities for object recognition

Tagging systems

Collective representations, language and tagging systems

Multimodal integration

Connectionist modeling of multimodal integration

Publications

Higy, B., & Bell, P. 2018. Few-shot learning with attention-based sequence-to-sequence models. http://arxiv.org/abs/1811.03519.

Higy, B., Mereta, A., Metta, G. & Badino., L. Speech Recognition for the ICub Platform. Frontiers in Robotics and AI 5 (2018).

Savran, A., Tavarone, R., Higy, B., Badino, L., & Bartolozzi., C. Energy and Computation Efficient Audio-Visual Voice Activity Detection Driven by Event-Cameras. In Proceedings of 13th IEEE International Conference on Automatic Face Gesture Recognition. Xi’an, China, 2018.

Higy, B., Ciliberto, C., Rosasco, L. & Natale., L. Combining Sensory Modalities and Exploratory Procedures to Improve Haptic Object Recognition in Robotics. In 2016 IEEE-RAS 16th International Conference on Humanoid Robots (Humanoids), 117–24, Cancun, Mexico, 2016. https://doi.org/10.1109/HUMANOIDS.2016.7803263.

Contact

If you want to reach to me...

Contact details

Bertrand Higy
bertrand.higy@gmail.com
iCub Facility
Istituto Italiano di Tecnologia
Via Morego, 30
Genova 16163, Italy

On the web