Robot Audition

the process of endowing a robot with the capability to be audio aware

Audio Localization

Robots model of audition extracts egocentric location and through bayesian model integrates knowledge in space and time

Auditory attention

Cognitive model that enable the robots to attend salient source of sound in the environment

Deep Learning for Sound Localization and Classification

The deep network provided in this research proved to be optimal in resolving simultaneously localization and classification of sound sources in auditory landscape

Bayesian inference about temporal structures observed and generated by ego-motion

The deep network provided in this research proved to be optimal in resolving simultaneously localization and classification of sound sources in auditory landscape

Predictive Coding and its deep computational representation

The computational models of Bayesian Inference for predictive coding rely on deep neural networks DNN.

Video: Saliency based sensor fusion

The humanoid robot iCub attends the interaction thanks to multisensory attention system that integrates audio and visual saliency maps.

Video: Auditory attention enhanced by optimal head motor control for a humanoid robot iCub

The humanoid robot iCub perceives auditory landscape with the use of microphones in the head. The iCub infers the location of the speakers and optimises the estimation with specific head motor actions.

Publication Selection

Main publications in this research field resulting from state-of-the-art research and technological breakthroughs

2020 Audiovisual cognitive architecture for autonomous learning of face localisation by a Humanoid Robot

Gonzalez Billandon J., Sciutti A., Tata M., Sandini G., Rea F., International Conference on Robotics and Automation, ICRA2020

2019 A Bayesian System for Noise-Robust Binaural Sound Localisation for Humanoid Robots

Kothig A., Ilievski M., Grasse L., Rea F., Tata M. 2019, in IEEE International Symposium on Robotic and Sensors Environments, ROSE

2019 Cognitive Architecture for Joint Attentional Learning of word-object mapping with a Humanoid Robot

Gonzalez Billandon J., Grasse L., Sciutti A., Tata M., Rea F. 2019, in IROS 2019 - Workshop Deep Probabilistic Generative Models for Cognitive Architecture in Robotics

2017 A Binaural Beamforming Approach to Resolve Complex Auditory Scenes for Humanoid Robots

Ilievski M., Rea F., Sandini G., Tata M. In 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems

Collaboration in the specific research

Collaboration activated thanks to the research carried on in the field

University of Leftbridge

The Department of Neuroscience / Canadian Centre for Behavioural Neuroscience (CCBN)

Osaka University

Symbiotic Intelligent Systems Research Center at the Institute for Open and Transdisciplinary Research Initiatives

Essex University

School of Computer Science and Electronic Engineering (CSEE)

Specific research activities in the topic

The progress beyond the state-of-the-art in the specific research activities carried on in this reseach topic.

Audio Localization

85%

Auditory attention

95%

Optimal Motor Control for Auditory Localization

80%

Deep Learning for Sound Localization and Classification

85%

Personal Activities: Dissemination

Teaching

Human Robot Interaction Course

The course is aimed at providing a clear understanding of what are the good features and limitations of the robotic platforms currently available. The participants will learn how to design and implement robot perceptual, motor abilities structured in a cognitive framework for natural human-robot interaction.

Conferences

List of attended conferences where my current research has been communicated and disseminated.

Invited Talks

The list of invitations for which I have been asked to present my results and to give my overview as expert in the field.

Social Events

Social events where the results of my reseach have been disseminated to broad public audience.

Products and Draft Patents

AuditoryAI

The invention is an new process to register the existinance of sound objects,localize them, prioritize them, and direct the behaviour of an agent (e.g. a robot) towards them.

gazeTracking

The product endows intelligent system with one camera to infer the gaze direction by looking at the head orientation and the iris position.

-----------------------

------------------------

-------------------------

Testimonials

-------------------------------------------------

Prof. Samia Nefti Meziani/Robotics and Automation, Salford University

--------------------------------------------------

Prof. Davide Brugali/ Robotics, Universita degli Studi di Bergamo

Partners

In the quest of enabling function robots in our lives, trustful Partners shared the road to impacting solutions.

Phone

t:+39 010 71781420

Personal

STAY TUNED

Activities

Partners