Edoardo Ragusa
Papers
9
Total Citations
56
H-Index
5
About
Edoardo Ragusa is a researcher specializing in embedded machine learning, computer vision, and human-robot interaction, with a particular focus on making deep learning practical for resource-constrained and wearable robotic systems. His most significant contributions center on affordance detection and segmentation — the task of identifying how objects can be interacted with — and adapting these capabilities for deployment on low-power, portable hardware where conventional deep learning approaches are computationally prohibitive. Ragusa has pioneered hardware-aware approaches to neural network design, including novel neural architecture search (HW NAS) methods capable of running on devices with as little as 512MB of RAM, producing efficient convolutional neural networks tailored to embedded platforms. His work spans RGB and RGB-D sensing modalities, grasping classification from video, and container mass estimation, reflecting a broad commitment to practical perception systems for assistive and wearable robotics. His most cited work, "Hardware-Aware Affordance Detection for Application in Portable Embedded Systems" (2021, 16 citations), exemplifies his core thesis: that sophisticated perception need not demand high-end computing. With a growing body of publications accumulating over 50 citations collectively, Ragusa is establishing himself as a thoughtful contributor at the intersection of efficient deep learning and real-world robotic sensing applications.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Data-Driven Video Grasping Classification for Low-Power Embedded System6 citations · 2019
- 5Container Localisation and Mass Estimation with an RGB-D Camera6 citations · 2022
- 6An Affordance Detection Pipeline for Resource-Constrained Devices5 citations · 2021
- 7
- 8Video Grasping Classification Enhanced with Automatic Annotations2 citations · 2021
- 9