Marco Cannici

Postdoctoral Researcher

Smart Eyewear Lab, Politecnico di Milano and EssilorLuxottica

Biography

I am a Postdoctoral Researcher at the Smart Eyewear Lab (Politecnico di Milano & EssilorLuxottica), working on event-based vision for intelligent wearable devices, with a particular focus on eye tracking and embedded perception systems. Previously, I was part of the Robotics and Perception Group, led by Prof. Davide Scaramuzza, where I worked on advancing visual perception with event-based cameras. Throughout my career, I have contributed to advancements in neural radiance field reconstruction under fast motion, low-power perception with efficient neural networks and neuromorphic hardware, and egocentric and monocular vision for real-time applications. My work increasingly focuses on bridging research and deployment, including the design of real-time perception pipelines and on-device inference systems for resource-constrained platforms. In the past, I have also explored integrating differentiable algorithms into end-to-end trainable models, with a focus on enhancing their robustness and explainability.

Interests

Eye Tracking
Egocentric Perception
Event-Based Cameras
Representation Learning
Deep Learning
Computer Vision

Education

Postdoc in Event-Based Cameras and Computer Vision, 2022-2025

Robotics and Perception Group, University of Zurich
PhD Student in Information Technology, 2018-2022

Politecnico di Milano
MSc in Computer Science and Engineering, 2018

Politecnico di Milano
BSc in Computer Science and Engineering, 2015

Politecnico di Milano

Publications

Quickly discover relevant content by filtering publications.

Mariam Hassan, Sebastian Stapf, Ahmad Rahimi, Pedro Rezende, Yasaman Haghighi, David Brüggemann, Isinsu Katircioglu, Lin Zhang, Xiaoran Chen, Suman Saha, others (2025). GEM: A Generalizable Ego-Vision Multimodal World Model for Fine-Grained Ego-Motion, Object Dynamics, and Scene Composition Control. arXiv preprint arXiv:2412.11198.

Cite

Roberto Pellerito, Marco Cannici, Daniel Gehrig, Joris Belhadj, Olivier Dubois-Matra, Massimo Casasco, Davide Scaramuzza (2024). Deep Visual Odometry with Events and Frames. 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

Cite DOI

Riccardo Santambrogio, Marco Cannici, Matteo Matteucci (2024). FARSE-CNN: Fully Asynchronous, Recurrent and Sparse Event-Based CNN. European Conference on Computer Vision.

Cite DOI

Vincenzo Polizzi, Marco Cannici, Davide Scaramuzza, Jonathan Kelly (2024). FaVoR: Features via Voxel Rendering for Camera Relocalization. arXiv preprint arXiv:2409.07571.

Cite

Caterina Caccavella, Federico Paredes-Vallés, Marco Cannici, Lyes Khacef (2024). Low-power event-based face detection with asynchronous neuromorphic hardware. 2024 International Joint Conference on Neural Networks (IJCNN).

Cite DOI

Marco Cannici, Davide Scaramuzza (2024). Mitigating motion blur in neural radiance fields with events and frames. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.

Cite DOI

Anish Bhattacharya, Marco Cannici, Nishanth Rao, Yuezhan Tao, Vijay Kumar, Nikolai Matni, Davide Scaramuzza (2024). Monocular Event-Based Vision for Obstacle Avoidance with a Quadrotor. arXiv preprint arXiv:2411.03303.

Cite DOI

Yifei Liu, Mathias Gehrig, Nico Messikommer, Marco Cannici, Davide Scaramuzza (2024). Revisiting token pruning for object detection and instance segmentation. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision.

Cite DOI

Ling Gao, Hang Su, Daniel Gehrig, Marco Cannici, Davide Scaramuzza, Laurent Kneip (2023). A 5-point minimal solver for event camera relative motion estimation. Proceedings of the IEEE/CVF International Conference on Computer Vision.

Cite DOI

Pierpaolo Mancini, Marco Cannici, Matteo Matteucci (2023). Deep learning for asteroids autonomous terrain relative navigation. Advances in Space Research.

Cite DOI

Francesco Lattari, Marco Cannici, Gianluca Murdaca, Alessio Rucci, Claudio Maria Prati, Matteo Matteucci (2023). NeuralPUMA: Learning to Phase Unwrap Through Differentiable Graph Cuts. IEEE Transactions on Signal Processing.

Cite DOI

Alessio Bonfiglio, Marco Cannici, Matteo Matteucci (2023). SoftCut: A Fully Differentiable Relaxed Graph Cut Approach for Deep Learning Image Segmentation. International Conference on Machine Learning, Optimization, and Data Science.

Cite DOI

Andrea Simpsi, Marco Roggerini, Marco Cannici, Matteo Matteucci (2022). 6 DoF Pose Regression via Differentiable Rendering. International Conference on Image Analysis and Processing.

Cite DOI

Chiara Plizzari, Mirco Planamente, Gabriele Goletto, Marco Cannici, Emanuele Gusso, Matteo Matteucci, Barbara Caputo (2022). E2 (go) motion: Motion augmented event stream for egocentric action recognition. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition.

Cite

Mirco Planamente, Chiara Plizzari, Marco Cannici, Marco Ciccone, Francesco Strada, Andrea Bottino, Matteo Matteucci, Barbara Caputo (2021). DA4Event: Towards Bridging the Sim-to-Real Gap for Event Cameras Using Domain Adaptation. IEEE Robotics and Automation Letters.

PDF Cite Code DOI

Alberto Archetti, Marco Cannici, Matteo Matteucci (2021). Neural Weighted A*: Learning Graph Costs and Heuristics with Differentiable Anytime A*. International Conference on Machine Learning, Optimization, and Data Science.

PDF Cite Code Dataset

Chiara Plizzari, Marco Cannici, Matteo Matteucci (2021). Skeleton-based action recognition via spatial and temporal transformer networks. Computer Vision and Image Understanding.

PDF Cite Code DOI

Marco Cannici, Chiara Plizzari, Mirco Planamente, Marco Ciccone, Andrea Bottino, Barbara Caputo, Matteo Matteucci (2021). N-ROD: a Neuromorphic Dataset for Synthetic-to-Real Domain Adaptation. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).

PDF Cite Code Dataset Project Video DOI

Alberto Archetti, Marco Cannici, Matteo Matteucci (2021). Neural weighted a*: Learning graph costs and heuristics with differentiable anytime a. International Conference on Machine Learning, Optimization, and Data Science.

Cite DOI

Chiara Plizzari, Marco Cannici, Matteo Matteucci (2021). Spatial Temporal Transformer Network for Skeleton-Based Action Recognition. Pattern Recognition. ICPR International Workshops and Challenges.

PDF Cite Code DOI

Marco Cannici, Marco Ciccone, Andrea Romanoni, Matteo Matteucci (2020). A Differentiable Recurrent Surface for Asynchronous Event-Based Data. Computer Vision – ECCV 2020.

Cite DOI

Marco Cannici, Marco Ciccone, Andrea Romanoni, Matteo Matteucci (2020). A Differentiable Recurrent Surface for Asynchronous Event-Based Data. Computer Vision – ECCV 2020.

PDF Cite Code Project Video DOI

Marco Cannici, Marco Ciccone, Andrea Romanoni, Matteo Matteucci (2019). Asynchronous Convolutional Networks for Object Detection in Neuromorphic Cameras. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).

PDF Cite Code Video DOI

Marco Cannici, Marco Ciccone, Andrea Romanoni, Matteo Matteucci (2019). Attention Mechanisms for Object Recognition With Event-Based Cameras. 2019 IEEE Winter Conference on Applications of Computer Vision (WACV).

PDF Cite DOI