Papers
13
Total Citations
258
H-Index
9
About
Giulia Pasquale is a robotics and computer vision researcher whose work sits at the intersection of deep learning and autonomous robotic perception. She is best known for her pioneering efforts in equipping humanoid robots — particularly the iCub platform — with robust visual recognition capabilities, addressing one of the most persistent challenges in deploying robots in real-world, unstructured environments. Her most cited work, "Object identification from few examples by improving the invariance of a Deep Convolutional Neural Network" (2016, 50 citations), exemplifies her focus on making object recognition systems both data-efficient and generalizable — a critical need in robotics where large annotated datasets are costly to obtain. This theme runs throughout her research, from early explorations of off-the-shelf deep convolutional networks for robot perception (2015) to interactive and weakly supervised data collection strategies that reduce labeling burden. Her 2018 survey, "Are we done with object recognition?" (46 citations), reflects her broader ambition to critically assess progress in the field. Pasquale has also contributed meaningfully to object detection, depth-driven visual attention, superquadric-based grasping, and kernel-based segmentation methods, accumulating over 247 citations. Her body of work represents a sustained, rigorous effort to bridge cutting-edge computer vision research with the practical demands of humanoid robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2Are we done with object recognition? The iCub robot’s perspective46 citations · 2018
- 3Teaching iCub to recognize objects using deep convolutional neural networks35 citations · 2015
- 4
- 5On-line object detection: a robotics challenge26 citations · 2019
- 6
- 7Improving Superquadric Modeling and Grasping with Prior on Object Shapes13 citations · 2018
- 8
- 9
- 10Fast Object Segmentation Learning with Kernel-based Methods for Robotics8 citations · 2021