Fabio Tosi, PhD

Post-Doc Researcher @ Univeristy of Bologna
fabio.tosi5@unibo.it
CV

NEWS:
2021:
- April - 3 papers accepted to TPAMI.
- February - 1 paper accepted to CVPR 2021 with Prof. Andreas Geiger.
- February - New position as research fellow at the University of Bologna for a project funded by Huawei Technologies Co. Ltd.
2020:
- October - I was acknowledged as Outstanding Reviewer at ECCV 2020!
- July - 2 papers accepted to ECCV!
- June - CVPR Tutorial: Learning and understanding single image depth estimation in the wild with Matteo, Filippo, Stefano, Godard Clément, Watson Jamie, Firman Michael, and Brostow Gabriel J.!
- April - Visiting PhD student at the Max Planck Institute for Intelligent System and University of Tübingen ( Autonomous Vision Group ) - Supervisor: Prof. Andreas Geiger
- February - 2 papers accepted to CVPR
- January - 1 paper accepted to CVIU

Short bio

I received my Bachelor degree in 2014 ("Refinement techniques for depth data generated by a stereo vision system", advisor: Prof. Stefano Mattoccia) and my Master degree in 2017 ("Confidence measures and depth map refinement algorithm", advisor: Prof. Stefano Mattoccia, summa cum laude) at Alma Mater Studiorum, University of Bologna.
I received my PhD degree in 2020 at Department of Computer Science and Engineering (DISI). Currently, I'm a research fellow at the University of Bologna for a project funded by Huawei Technologies Co. Ltd.

Research interests

3D sensing and applications, embedded vision, machine learning, deep learning

My team

Prof. Stefano Mattoccia
Matteo Poggi (Post-Doc)
Filippo Aleotti (PhD)

Code

Github

Tutorials and demos Experience as reviewer
  • IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
  • IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • European Conference on Computer Vision (ECCV)
  • IEEE International Conference on Computer Vision (ICCV)
  • CVPR 2019 - 3nd International Workshop on Computer Vision for UAVs (UAVision2019)
  • ECCV 2018 - 2nd International Workshop on Computer Vision for UAVs (UAVision2018)
  • IEEE Transactions on Neural Networks and Learning Systems (TNNLS)
  • International Journal of Distributed Sensor Networks (IJDSN)
  • IEEE Transactions on Image Processing (TIP)
  • Journal of Electronic Imaging (JEI)
Acknowledgement
  • Outstanding Reviewer, ECCV 2020. (PDF)
Patents
  • Method for determining the depth from a single image and system thereof - M.Poggi, F. Aleotti, Fabio Tosi, S. Mattoccia, V. Peluso, A. Cipolletta, A. Calimera (pending)
  • Method for determining the confidence of a disparity map through a self-adaptive learning of a neural network, and sensor system thereof (pending)
  • Depth determination method based on images, self-adaptive neural networks, and relative system (Real-Time Self Adaptive Deep Stereo) (pending)
  • Depth determination method based on images, and relative system (pending)
Pubblications

Google Scholar

International Conferences and Workshops

  • F. Tosi, Y. Liao, C. Schmitt, A. Geiger, "SMD-Nets: Stereo Mixture Density Networks", accepted at The IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2021), June 19-25, 2021, Nashville, Tennessee, US.
  • F. Aleotti, F. Tosi, L. Zhang, M. Poggi, S. Mattoccia, "Reversing the cycle: self-supervised deep stereo through enhanced monocular distillation", accepted at The European Conference on Computer Vision (ECCV 2021), August 23-28, 2020, Glasgow, UK. (PDF) (Code)
  • M. Poggi, F. Aleotti, F. Tosi, G. Zaccaroni, S. Mattoccia, "Self-adapting confidence estimation for stereo", accepted at The European Conference on Computer Vision (ECCV 2021), August 23-28, 2020, Glasgow, UK. (PDF) (Code)
  • F. Tosi, F. Aleotti, P. Zama Ramirez, M. Poggi, S. Salti, L. Di Stefano and S. Mattoccia, "Distilled Semantics for Comprehensive Scene Understanding from Videos", accepted at The IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2020), June 16-18, 2020, Seattle, Washington, US. (PDF) (Code)
  • M. Poggi, F. Aleotti, F. Tosi and S. Mattoccia, "On the uncertainty of self-supervised monocular depth estimation", accepted at The IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2020), June 16-18, 2020, Seattle, Washington, US. (PDF) [CODE]
  • M. Poggi, F. Tosi, F. Aleotti, S. Mattoccia, "Leveraging a weakly adversarial paradigm for joint learning of disparity and confidence estimation", accepted at the International Conference on Pattern Recognition (ICPR 2020)
  • F. Aleotti, M. Poggi, F. Tosi, S. Mattoccia, "Learning end-to-end scene flow by distilling single tasks knowledge", accepted at the 34th AAAI Conference on Artificial Intelligence, New York, US, February 7-12, 2020 (PDF)
  • Matteo Poggi, Davide Pallotti, Fabio Tosi and Stefano Mattoccia, "Guided Stereo Matching", accepted at The IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2019), June 16-21, 2019, Long Beach, California, US. (PDF)(Demo Code)
  • Fabio Tosi, Filippo Aleotti, Matteo Poggi and Stefano Mattoccia, "Learning monocular depth estimation infusing traditional stereo knowledge", accepted at The IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2019), June 16-21, 2019, Long Beach, California, US. (PDF)(supplementary)(Code)
  • Alessio Tonioni, Fabio Tosi, Matteo Poggi, Stefano Mattoccia and Luigi Di Stefano, "Real-time self-adaptive deep stereo", accepted at The IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2019), June 16-21, 2019, Long Beach, California, US (ORAL) (PDF)(supplementary)(Code)
  • F. Tosi, M. Poggi, S. Mattoccia, "Leveraging confident points for accurate depth refinement on embedded systems", accepted at The IEEE Embedded Vision Workshop (EVW 2019), June 16, 2019, Long Beach, California, US. (PDF)
  • Valentino Peluso, Antonio Cipolletta, Andrea Calimera, Matteo Poggi, Fabio Tosi and Stefano Mattoccia, "Enabling Energy-Efficient Unsupervised Monocular Depth Estimation on ARMv7-Based Platforms", accepted at Design, Automation and Design in Europe (DATE 2019), March 29-29, 2019, Florence, Italy. (PDF)
  • M. Poggi, F. Tosi, S. Mattoccia, "Learning monocular depth estimation with unsupervised trinocular assumptions", accepted at The 6th international conference on 3D Vision (3DV 2018), September 5-8, 2018, Verona, Italy. (PDF)(Code)(Video)
  • P. Zama Ramirez, M. Poggi, F. Tosi, S. Mattoccia, L. Di Stefano, "Geometry meets semantic for semi-supervised monocular depth estimation", accepted at 14th Asian Conference on Computer Vision (ACCV 2018), December 2-6, 2018, Perth, Australia (PDF)(Code)
  • F. Tosi, M. Poggi, A. Benincasa, S. Mattoccia, "Beyond local reasoning for stereo confidence estimation with deep learning", accepted at the 15th European Conference on Computer Vision (ECCV 2018), September 8-14, 2018, Munich, Germany. (PDF)(Code)
  • M. Poggi, F. Aleotti, F. Tosi, S. Mattoccia, "Towards real-time unsupervised monocular depth estimation on CPU", accepted at IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2018), October 1-5, 2018, Madrid, Spain. (PDF)(Code)(Video)
  • F. Aleotti, F. Tosi, M. Poggi, S. Mattoccia, "Generative Adversarial Networks for unsupervised monocular depth prediction", accepted at 3D Reconstruction in the Wild 2018 (3DRW2018), in conjunction with (ECCV 2018), Munich, Germany, September 14, 2018. (PDF)
  • M. Poggi, F. Tosi, S. Mattoccia, "Quantitative evaluation of confidence measures in a machine learning world", accepted at The IEEE International Conference on Computer Vision (ICCV 2017), October 22-29, 2017, Venezia, Italy SPOTLIGHT (PDF)
  • F. Tosi, M. Poggi, A.Tonioni, L. Di Stefano, S. Mattoccia, "Learning confidence measures in the wild", accepted at The 28th British Machine Vision Conference (BMVC 2017), September 5-7, 2017, London, UK (PDF)(Code)
  • M. Poggi, F. Tosi, S. Mattoccia, "Efficient confidence measures for embedded stereo", accepted at The 19th International Conference on Image Analysis and Processing (ICIAP 2017), September 11-15, 2017, Catania, Italy (PDF)
  • M. Poggi, F. Tosi, S. Mattoccia, "Even More Confident predictions with deep machine-learning", accepted at The IEEE Embedded Vision Workshop (EVW 2017), July 21, 2017, Honolulu, Hawaii, US (PDF)

Journals
  • M. Poggi, F. Tosi, K. Batsos, P. Mordohai, S. Mattoccia, "On the Synergies between Machine Learning and Binocular Stereo for Depth Estimation from Images: a Survey", IEEE Transaction on Pattern Analysis and Machine Intelligence (TPAMI)
  • M. Poggi, S. Kim, F. Tosi, S. Kim, F. Aleotti, D. Min, K. Sohn, and S. Mattoccia, "On the Confidence of Stereo Matching in a Deep-Learning Era: a Quantitative Evaluation", IEEE Transaction on Pattern Analysis and Machine Intelligence (TPAMI)
  • V. Peluso, A. Cipolletta, A. Calimera, M. Poggi, F. Tosi, F. Aleotti, S. Mattoccia, “Monocular Depth Perception on Microcontrollers”, accepted on IEEE Transactions on Circuits and Systems for Video Technology (TCSVT)
  • F. Aleotti, G. Zaccaroni, L. Bartolomei, M. Poggi, F. Tosi, S. Mattoccia, “Real-time singleimage depth perception in the wild with handheld devices", MDPI Sensors
  • Tosi, M. Rocca, F. Aleotti, M. Poggi, S. Mattoccia, F. Tauro, E. Toth, S.Grimaldi, "Enabling image-based streamflow monitoring at the edge", MDPI Remote Sensing, RemoteSensing
  • M. Poggi, F. Tosi, S. Mattoccia, "Good cues to learn from scratch a confidence measurefor passive depth sensors", IEEE Sensors Journal, Sensors
  • M. Poggi, F. Tosi, S. Mattoccia, "Learning a confidence measure in the disparity domain from O(1) features", Computer Vision and Image Understanding (CVIU)
  • M. Poggi, G. Agresti, F. Tosi, P. Zanuttigh, S. Mattoccia, "Confidence Estimation for ToF and Stereo Sensorsand its Application to Depth Data Fusion", IEEE Sensors
  • F. Tauro, F. Tosi, S. Mattoccia, E. Toth, R. Piscopia, S. Grimaldi, "Optical Tracking Velocimetry (OTV): leveraging optical flow and trajectory-based filtering for surface streamflow observations", Remote Sensing , 2018, 10(12), 2010 (PDF)
Pre-print
  • M. Poggi, A. Tonioni,F. Tosi, L. Di Stefano, S. Mattoccia, “Continual Adaptation for DeepStereo", at the IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)