Junior Assistant Professor · University of Bologna
Fabio Tosi, PhD
I work at the intersection of computer vision and deep learning, on machines that understand the 3D structure of the world — stereo matching, monocular depth estimation, neural rendering and SLAM.
Department of Computer Science and Engineering (DISI), University of Bologna. I am also interested in making these models small and fast enough to run on resource-constrained devices.

Bologna, Italy · fabio.tosi5@unibo.it
About
- Stereo Matching
- Monocular Depth Estimation
- Neural Rendering
- SLAM
- Efficient Deep Learning
I received my PhD in 2021 from the University of Bologna, supervised by Professor Stefano Mattoccia, and my Master’s (2017) and Bachelor’s (2014) degrees in Computer Engineering from the same university.
In 2020, I was a visiting PhD student in the Autonomous Vision Group (AVG) led by Professor Andreas Geiger at the Max Planck Institute for Intelligent Systems and the University of Tübingen. In 2022, I received the Best PhD Thesis Award from the Italian Association for Research in Computer Vision, Pattern Recognition and Machine Learning (CVPL).
Since 2024, I serve as Associate Editor for Pattern Recognition (Elsevier). I have also served as Area Chair at CVPR 2026, ECCV 2026, ACCV 2026, and ICIAP 2025, and as Associate Editor at IROS 2025.
I teach Accelerated Computing Systems (CUDA C, GPU architectures, high-performance computing) in the Master’s program in Computer Engineering, and GPU-accelerated computing for AI at the PhD level. I also teach Fundamentals of Computer Science (logic networks, computer architectures) in the Mechatronics program.
News
- 09/2026 Upcoming: I am co-organizing the 3rd Workshop on Neural SLAM (NeuSLAM), held in conjunction with ECCV 2026 in Malmö, Sweden — September 8, afternoon.
- 06/2026 3 papers accepted to ECCV 2026, 1 paper accepted to IROS 2026!
- 05/2026 Honored to be recognized as an Outstanding Area Chair at CVPR 2026!
- 02/2026 3 papers at CVPR 2026, 1 at CVPR Findings 2026!
- 02/2026 Our survey on NeRF & 3DGS-based SLAM accepted to T-RO!
- 12/2025 2 paper accepted to WACV 2025 and AAAI 2025 (Oral)!
- 09/2024 Grateful for the Outstanding Reviewer recognition at ICCV 2025!
- 09/2025 1 paper accepted with Rawmantic AI to NIPS 2025!
- 06/2025 2 paper accepted to ICCV 2025!
- 06/2025 One paper is set for IROS 2025, and another for IJCV!
- 05/2025 Grateful for the recognition as an Outstanding Reviewer at CVPR 2025!
Earlier newsHide earlier news
- 02/2025 Our paper “Stereo Anywhere: Robust Zero-Shot Deep Stereo Matching Even Where Either Stereo or Mono Fail” has been accepted at CVPR 2025!
- 01/2025 1 paper accepted to ICRA 2025!
- 12/2024 Our survey paper, “A Survey on Deep Stereo Matching in the Twenties”, has been accepted at the IJCV journal!
- 11/2024 Our research with Sadra and Fatma received the Best Poster Award at BMVC 2024!
- 07/2024 1 paper accepted to WACV 2024!
- 09/2024 Grateful for the Outstanding Reviewer recognition at ECCV 2024!
- 07/2024 Our paper “Diffusion Models for Monocular Depth Estimation” has been accepted at ECCV 2024!
- 06/2024 Our CVPR tutorial on deep stereo matching is now available! Get all the insights here!
- 06/2024 Our extended version of Neural Disparity Refinement has been accepted for publication in the TPAMI journal!
- 05/2024 Honored to be recognized as a Outstanding Reviewer at CVPR 2024!
- 10/2023 I am proud to announce my new role as a Junior Assistant Professor (RTDA) at the Department of Computer Science and Engineering (DISI)!
- 05/2023 It is with great pleasure that I announce my achievement as an Outstanding Reviewer at CVPR 2023! (Award Certificate)
- 02/2023 I received my National Scientific Habilitation (09/H1)
- 11/2022 🏆 Celebrating the recognition of the Best PhD Thesis Award from the Italian Association for Computer Vision Research (CVPL 2022)
- 09/2021 Best Paper Honorable Mention to our work “Neural Disparity Refinement for Arbitrary Resolution Stereo”
Selected publications
A selection — the full list is on Google Scholar · * indicates joint first authorship

ZipDepth: Bringing Lightweight Zero-Shot Monocular Depth Anywhere, on Any Device New
ECCV|A++|European Conference on Computer Vision

DINO-SLAM: DINO-informed RGB-D SLAM for Neural Implicit and Explicit Representations New
ECCV|A++|European Conference on Computer Vision

MAGiSt3R: Multi-Agent Feed-forward 3D Reconstruction from Monocular RGB Videos New
ECCV|A++|European Conference on Computer Vision

Bidirectional Cross-Modal Prompting for Event-Frame Asymmetric Stereo
CVPR|A++|Conference on Computer Vision and Pattern Recognition

EventHub: Data Factory for Generalizable Event-Based Stereo Networks without Active Sensors
CVPR|A++|Conference on Computer Vision and Pattern Recognition

StereoSpace: Depth-Free Synthesis of Stereo Geometry via End-to-End Diffusion in a Canonical Space
CVPR Findings|Conference on Computer Vision and Pattern Recognition – Findings

Ov3R: Open-Vocabulary Semantic 3D Reconstruction from RGB Videos
CVPR|A++|Conference on Computer Vision and Pattern Recognition

How NeRFs and 3D Gaussian Splatting are Reshaping SLAM: a Survey
T-RO|Q1|IEEE Transactions on Robotics|vol. 42, pp. 1405–1427

FoundationSLAM: Unleashing the Power of Depth Foundation Models for End-to-End Dense Visual SLAM Oral
AAAI|A++|AAAI Conference on Artificial Intelligence

WarpRF: Multi-View Consistency for Training-Free Uncertainty Quantification and Applications in Radiance Fields
WACV|A|IEEE/CVF Winter Conference on Applications of Computer Vision

Eve3D: Elevating Vision Models for Enhanced 3D Surface Reconstruction via Gaussian Splatting
NeurIPS|A++|Conference on Neural Information Processing Systems

FlowSeek: Optical Flow Made Easier with Depth Foundation Models and Motion Bases
ICCV|A++|International Conference on Computer Vision

A Survey on Deep Stereo Matching in the Twenties
IJCV|Q1|International Journal of Computer Vision

Stereo Anywhere: Robust Zero-Shot Deep Stereo Matching Even Where Either Stereo or Mono Fail
CVPR|A++|Conference on Computer Vision and Pattern Recognition

Self-Evolving Depth-Supervised 3D Gaussian Splatting from Rendered Stereo Pairs 🏆 Best Poster Award
BMVC|A|British Machine Vision Conference

Diffusion Models for Monocular Depth Estimation: Overcoming Challenging Conditions
ECCV|A++|European Conference on Computer Vision

Federated Online Adaptation for Deep Stereo
CVPR|A++|Conference on Computer Vision and Pattern Recognition

GO-SLAM: Global Optimization for Consistent 3D Instant Reconstruction
ICCV|A++|International Conference on Computer Vision

MonoViT: Self-supervised Monocular Depth Estimation with a Vision Transformer
3DV|A-|International Conference on 3D Vision

Open Challenges in Deep Stereo: the Booster Dataset
CVPR|A++|Conference on Computer Vision and Pattern Recognition

RGB-Multispectral Matching: Dataset, Learning Methodology, Evaluation
CVPR|A++|Conference on Computer Vision and Pattern Recognition

Neural Disparity Refinement for Arbitrary Resolution Stereo 🏆 Best Paper Honorable Mention
3DV|A-|International Conference on 3D Vision

On the Synergies Between Machine Learning and Binocular Stereo for Depth Estimation From Images: A Survey
TPAMI|Q1|IEEE Transactions on Pattern Analysis and Machine Intelligence

SMD-Nets: Stereo Mixture Density Networks
CVPR|A++|Conference on Computer Vision and Pattern Recognition

Distilled Semantics for Comprehensive Scene Understanding from Videos
CVPR|A++|Conference on Computer Vision and Pattern Recognition

Reversing the Cycle: Self-Supervised Deep Stereo through Enhanced Monocular Distillation
ECCV|A++|European Conference on Computer Vision

Learning Monocular Depth Estimation Infusing Traditional Stereo Knowledge
CVPR|A++|Conference on Computer Vision and Pattern Recognition

Real-Time Self-Adaptive Deep Stereo Oral
CVPR|A++|Conference on Computer Vision and Pattern Recognition









