Papers
24
Total Citations
404
H-Index
11
About
Francesca Odone is a prominent computer vision and robotics researcher whose work sits at the intersection of visual recognition, human-robot interaction, and machine learning. Her research has made substantial contributions to enabling humanoid robots — particularly the iCub platform — to perceive, understand, and interact with the world more naturally and reliably. Odone's most cited work, "Keep It Simple and Sparse: Real-Time Action Recognition" (2017, 81 citations), demonstrates her commitment to computationally efficient yet powerful recognition systems. Across her career, she has tackled core challenges in robotic perception, including object recognition using deep convolutional neural networks, eye-hand coordination, stereo estimation, and biological motion detection — skills she argues are foundational to safe, intuitive human-robot collaboration. Her iCub-focused studies, including the widely referenced iCub World dataset paper, have helped establish benchmark resources that benefit the broader robotics community. Odone also advances multimodal understanding through contributions like the MoCA dataset, capturing fine-grained cooking actions via motion capture and multi-view video. With over 300 combined citations across her most recognized works, her research meaningfully shapes how autonomous agents learn to see, act, and cooperate alongside humans in real-world environments.
Research Focus
Key Achievements
Top Papers
- 1Keep It Simple and Sparse: Real-Time Action Recognition81 citations · 2017
- 2Are we done with object recognition? The iCub robot’s perspective46 citations · 2018
- 3
- 4Teaching iCub to recognize objects using deep convolutional neural networks35 citations · 2015
- 5
- 6iCub World: Friendly Robots Help Building Good Vision Data-Sets30 citations · 2013
- 7
- 8Ask the Image: Supervised Pooling to Preserve Feature Locality16 citations · 2014
- 9
- 10Object segmentation using independent motion detection12 citations · 2015