Back to list

Giulia Pasquale

Senior Technician

Research Line

Humanoid Sensing and Perception

Center

IIT Central Research Labs Genova

Contacts

Via San Quirico 19D, 16126, Genova, IT
+39 010 2898 235
Contact Me

Social profiles

linkedin googlescholar github

About

I am a research engineer with three-year professional experience in robotic and machine vision systems. As evidenced by my achievements, my expertise includes machine learning, computer vision, robotics and parallel computing. My work aims at pushing forward the visual perception capabilities of automated systems (robots and mobile devices) to enable reliable and flexible real-world applications. I like tackling every development aspect, from the study and design, implementation and benchmark, to the deployment and optimization.

I received the B. Sc. Degree in Biomedical Engineering and the M. Sc. Degree in Bioengineering (both with Honours), respectively in 2010 and 2013, at the University of Genoa.
After a short research fellowship at the National Research Council of Italy (IMATI), from 2014 to 2017 I pursued the Ph. D. in Bioengineering and Robotics, at the Istituto Italiano di Tecnologia in collaboration with the University of Genoa, under the supervision of Dr. Lorenzo Natale and Prof. Lorenzo Rosasco.

Since 2017 I am a Postdoctoral Researcher at the Istituto Italiano di Tecnologia, Humanoid Sensing and Perception (HSP) research line. I am also affiliated with the Laboratory for Computational and Statistical Learning (LCSL), University of Genoa, DIBRIS.

My research currently focuses on automatic visual object recognition and localization, including scientific and technological aspects, ranging from data collection and annotation, to robustness and reliability under varying conditions, ultimately targeting efficient and fast learning applications.

Interests

machine learning deep learning computer vision robotics parallel computing

Projects

iCub World Datasets

A large data collection project (website) that aims at benchmarking visual object recognition and localization methods deployed to humanoid robots like the iCub.

As part of my PhD research activity, I substantially contributed to this project by collecting two dataset releases, one of which (iCubWorld Transformations Dataset) is the largest (> 400K images representing 200 objects into 20 categories) and the latest at the moment.

Groups involved: HSP and iCub research lines at the IIT; LCSL lab and SlipGURU research unit at the University of Genoa, DIBRIS.

On-the-fly Object Recognition

A deep learning-based visual object recognition pipeline for humanoid robots.

As part of my PhD research activity, I investigated and deployed deep learning methods to the on-the-fly object recognition pipelines used on the iCub and R1 robots.

The system (and its evolutions) enable quick learning of novel objects with natural user interaction through seamless data collection and fast training (video on iCub, video on R1). This system was exploited, as an example, for the grasping of novel objects (video) and learning of tool affordances (video); it was also deployed to other platforms, like the WalkMan robot, and adopted by the team winner of the Kuka Innovation Award 2018.

Groups involved: HSP and iCub research lines at the IIT; LCSL lab and SlipGURU research unit at the University of Genoa, DIBRIS.

On-the-fly Object Detection

As part of my Postdoctoral research activity, I am collaborating to the extension of the recognition systems used on the iCub and R1 robots to use deep learning methods for detecting, individually localizing and recognizing multiple objects in the scene (video).

Groups involved: HSP and iCub research lines at the IIT; LCSL lab at the University of Genoa, DIBRIS.

On-the-fly Category Learning from the Web

As part of my Postdoctoral research activity, I am collaborating to the extension of the recognition systems used on the iCub and R1 robots to learn novel object categories by gathering knowledge on-the-fly from the web (video).

Groups involved: HSP, iCub and Visual and Multimodal Applied Learning research lines at the IIT.

Awards

2019: Awarded of the Innovator Under 35 (TR35) Italy price by MIT Technology Review.

2014 and 2017: Donation of GPU cards via application to the NVIDIA GPU Grant Program.

2008: Awarded among the best ten students of the Scuola Politecnica, University of Genoa, for the academic year 2007/2008.

INFORMATION NOTICE ON COOKIES

IIT's website uses the following types of cookies: browsing/session, analytics, functional and third party cookies. Users can choose whether or not to accept the use of cookies and access the website. By clicking on "Further Information", the full information notice on the types of cookies will be displayed and you will be able to choose whether or not to accept them whilst browsing on the website
Further Information
Accept and close