Matteo Poggi

Ov3R. Python

ZipDepth. [ECCV 2026] Official implementation of "ZipDepth: Bringing Lightweight Zero-Shot Monocular Depth Anywhere, on Any Device". A compact 6.1M-parameter network for zero-shot monocular depth estimation, running in real time from server GPUs to mobile phones via knowledge distillation from foundation models.

Dataset3D. The first "ImageNet" 3D dataset.

Open-d4rt. Python

lightly-train. All-in-one training for vision models (YOLO, ViTs, RT-DETR, DINOv3): pretraining, fine-tuning, distillation.

lightly-studio. Curate, Annotate, and Manage Your Data in LightlyStudio.

vggt-omega. [CVPR 2026 Oral] VGGT Omega

Bi-CMPStereo. Bidirectional Cross-Modal Prompting for Event-Frame Asymmetric Stereo [CVPR 2026]

InfiniDepth. [CVPR 2026] InfiniDepth: Arbitrary-Resolution and Fine-Grained Depth Estimation with Neural Implicit Fields

eventhub. [CVPR 2026] EventHub: Data Factory for Generalizable Event-Based Stereo Networks without Active Sensors

StereoWorld. [CVPR 2026] Stereo World Model

Any-Resolution-Any-Geometry. Python

Velodepth. Video Depth Propagation [3DV 2026]

WAFT. [ICLR2026 - Oral] WAFT: Warping-Alone Field Transforms for Optical Flow

SparseSurf. [AAAI' 26]SparseSurf: Sparse-View 3D Gaussian Splatting for Surface Reconstruction

stereospace. Python

ReCoVEr. Removing Cost Volumes from Optical Flow Estimators (ICCV 2025 Oral)

GeoSVR. [NeurIPS'25 Spotlight] GeoSVR: Taming Sparse Voxels for Geometrically Accurate Surface Reconstruction

VocAlign. [BMVC'25] Official implementation of "Lost in Translation? Vocabulary Alignment for Source-Free Adaptation in Open-Vocabulary Semantic Segmentation"

Multimodal-SAM-Adapter. This repository contains download links to dataset, code snippets, and trained deep models of our work "Multimodal SAM-Adapter for Semantic Segmentation", by Iacopo Curti*, Pierluigi Zama Ramirez*, Alioscia Petrelli* , and Luigi Di Stefano*. * Equal Contribution University of Bologna

BridgeDepth. [ICCV 2025 Highlight] BridgeDepth: Bridging Monocular and Stereo Reasoning with Latent Alignment

depthanyevent. [ICCV 2025] Depth AnyEvent: A Cross-Modal Distillation Paradigm for Event-Based Monocular Depth Estimation

LFRD2. Code for Learnable Fractional Reaction-Diffusion Dynamics for Under-Display ToF Imaging and Beyond (ICCV 2025)

alltracker. AllTracker is a model for tracking all pixels in a video.

MultimodalStudio. Official repository of the CVPR paper "MultimodalStudio: A Heterogeneous Sensor Dataset and Framework for Neural Rendering across Multiple Imaging Modalities"

focoos. 🚀 Lightning-fast computer vision models. Fine-tune SOTA models with just a few lines of code. Ready for cloud ☁️ and edge 📱 deployment.

Flow-Anything. Official implementation of our paper "Flow-Anything: Learning Real-World Optical Flow Estimation from Large-Scale Single-view Images"

SemLA. [CVPR'25] Official implementation of "Semantic Library Adaptation: LoRA Retrieval and Fusion for Open-Vocabulary Semantic Segmentation"

alpha-NeuS. Python

stereoanywhere. [CVPR 2025] Stereo Anywhere: Robust Zero-Shot Deep Stereo Matching Even Where Either Stereo or Mono Fail

gaussian-opacity-fields. [SIGGRAPH Asia'24 & TOG] Gaussian Opacity Fields: Efficient Adaptive Surface Reconstruction in Unbounded Scenes

StereoGS. Source code for BMVC 2024 paper "Self-Evolving Depth-Supervised 3D Gaussian Splatting from Rendered Stereo Pairs"

MoGe. [CVPR'25 Oral] MoGe: Unlocking Accurate Monocular Geometry Estimation for Open-Domain Images with Optimal Training Supervision

LLaNA. Official code repository of LLaNA: Large Language and NeRF Assistant

depthsplat. [CVPR'25] DepthSplat: Connecting Gaussian Splatting and Depth

UnifiedGeneralization. Code for Self-Assessed Generation and CVPR2024 PAPER ADFACTORY

DepthAnyVideo. Depth Any Video with Scalable Synthetic Data (ICLR 2025)

depth-on-demand. ECCV24 Official Code for Depth on Demand: Streaming Dense Depth from a Low Frame-Rate Active Sensor

LayeredFlow. [ECCV 2024] LayeredFlow: A Real-World Benchmark for Non-Lambertian Multi-Layer Optical Flow

DepthCrafter. [CVPR 2025 Highlight] DepthCrafter: Generating Consistent Long Depth Sequences for Open-world Videos

map4d. Photo-realistic mapping of dynamic urban areas

drivestudio. A 3DGS framework for omni urban scene reconstruction and simulation.

eventvppstereo. [ECCV 2024] LiDAR-Event Stereo Fusion with Hallucinations

glomap. [DEPRECATED] GLOMAP - Global Structured-from-Motion Revisited

Diffusion4RobustDepth. [ECCV 2024] Diffusion Models for Monocular Depth Estimation: Overcoming Challenging Conditions

dataset.

ChronoDepth. ChronoDepth: Learning Temporally Consistent Video Depth from Video Diffusion Priors

awesome-implicit-representations. A curated list of resources on implicit neural representations.

SEA-RAFT. [ECCV2024 - Oral, Best Paper Award Candidate] SEA-RAFT: Simple, Efficient, Accurate RAFT for Optical Flow

fedstereo. Source code for "Federated Online Adaptation for Deep Stereo", CVPR 2024

ICML-2023-FedLAW. The is the official implementation of ICML 2023 paper "Revisiting Weighted Aggregation in Federated Learning with Neural Networks".

Awesome-Deep-Stereo-Matching. A curated list of awesome Deep Stereo Matching resources

TiO-Depth_pytorch. Python

dust3r. DUSt3R: Geometric 3D Vision Made Easy

Dataset. News: the 10k dataset is ready for download.

SuGaR. [CVPR 2024] Official PyTorch implementation of SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Rendering

FlowDiffusion_pytorch. Unofficial pytorch implementation of DDVM.

ramdepth. Official code for Range-Agnostic Multi-View Depth Estimation with Keyframe Selection (3DV 2024)

SpacetimeGaussians. [CVPR 2024] Spacetime Gaussian Feature Splatting for Real-Time Dynamic View Synthesis

FreeReg. [ICLR 2024] FreeReg: Image-to-Point Cloud Registration Leveraging Pretrained Diffusion Models and Monocular Depth Estimators

co-tracker. CoTracker is a model for tracking any point (pixel) on a video.

TouchSDF. Implementation of the DeepSDF paper

vppdc. [3DV 2024] Revisiting Depth Completion from a Stereo Matching Perspective for Cross-domain Generalization

Marigold. [CVPR 2024 - Oral, Best Paper Award Candidate] Marigold: Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation

FSGS. [ECCV 2024]"FSGS: Real-Time Few-Shot View Synthesis using Gaussian Splatting", Zehao Zhu*, Zhiwen Fan*, Yifan Jiang, Zhangyang Wang

GeoDream. GeoDream: Disentangling 2D and Geometric Priors for High-Fidelity and Consistent 3D Generation

Deformable-3D-Gaussians. [CVPR 2024] Official implementation of "Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction"

mip-splatting. [CVPR'24 Best Student Paper] Mip-Splatting: Alias-free 3D Gaussian Splatting

kitti-devkit. kitti-devkit for generating the error maps, KITTI-color-space disparity maps, and pfm2uint16png and uint16png2pfm converting

EmerNeRF. PyTorch Implementation of EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via Self-Supervision

READ. AAAI2023，implementation of "READ: Large-Scale Neural Scene Rendering for Autonomous Driving", the experimental results are significantly better than Nerf-based methods

waymax. A JAX-based simulator for autonomous driving research.

LRRU. Official implementation of ``LRRU: Long-short Range Recurrent Updating Networks for Depth Completion'', ICCV 2023.

omnimotion. Python

encord-active. The toolkit to test, validate, and evaluate your models and surface, curate, and prioritize the most valuable data for labeling.

Dynamic3DGaussians. Python

NeuRBF. Python

finite-solvability. Finite solvability repository - ICCV23, ECCV24, TPAMI26

gaussian-splatting. Original reference implementation of "3D Gaussian Splatting for Real-Time Radiance Field Rendering"

ICCV-2023-25-Papers. ICCV 2023-2025 Papers: Discover cutting-edge research from ICCV 2023-25, the leading computer vision conference. Stay updated on the latest in computer vision and deep learning, with code included. ⭐ support visual intelligence development!

vppstereo. [ICCV 2023] Active Stereo Without Pattern Projector; [IJCV] Active Stereo in the Wild through Virtual Pattern Projection

GO-SLAM. [ICCV2023] GO-SLAM: Global Optimization for Consistent 3D Instant Reconstruction

S-NeRF. [ICLR 2023 & TPAMI 2025] S-NeRF: Neural Radiance Fields for Street Views

DirectStereoRectification. "Rectifying Homographies for Stereo Vision: Analytical Solution for Minimal Distortion": algorithm to compute the optimal rectifying homographies that minimise perspective distortion.

neuralangelo. Official implementation of "Neuralangelo: High-Fidelity Neural Surface Reconstruction" (CVPR 2023)

differentiable_ransac. PyTorch Implementation of the ICCV 2023 paper: Generalized Differentiable RANSAC ($\nabla$-RANSAC).

zipnerf-pytorch. Unofficial implementation of ZipNeRF

CA-RANSAC. Python

slowtv_monodepth. Official repository for the ICCV2023 paper "Kick Back & Relax: Learning to Reconstruct the World by Watching SlowTV"

carla_garage. [ICCV'23] Hidden Biases of End-to-End Driving Models & A starter kit for the CARLA leaderboard 2.0.

hamlet. Source code for "To Adapt or Not to Adapt? Real-Time Adaptation for Semantic Segmentation", ICCV 2023

GasMono. Code for GasMono, accepted by ICCV 2023

TemporalStereo. [IROS 2023] TemporalStereo: Efficient Spatial-Temporal Stereo Matching Network

multi-object-segmentation. Code for "Multi-Object Discovery by Low-Dimensional Object Motion", ICCV 2023

sdfstudio. A Unified Framework for Surface Reconstruction

streetDownloader. Python

threestudio. A unified framework for 3D content generation.

Lite-Mono. [CVPR2023] Lite-Mono: A Lightweight CNN and Transformer Architecture for Self-Supervised Monocular Depth Estimation

transformers. 🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.

ppac_refinement. Probabilistic Pixel-Adaptive Refinement Networks (CVPR 2020)

inr2vec. [ICLR 2023] Deep Learning on Implicit Neural Representations of Shapes

pytorch-image-models. The largest collection of PyTorch image encoders / backbones. Including train, eval, inference, export scripts, and pretrained weights -- ResNet, ResNeXT, EfficientNet, NFNet, Vision Transformer (ViT), MobileNetV4, MobileNet-V3 & V2, RegNet, DPN, CSPNet, Swin Transformer, MaxViT, CoAtNet, ConvNeXt, and more

NeRF-Supervised-Deep-Stereo. A novel paradigm for collecting and generating stereo training data using neural rendering

CompletionFormer. [CVPR2023] CompletionFormer: Depth Completion with Convolutions and Vision Transformers

difflogic. A Library for Differentiable Logic Gate Networks

TurboNeRF. A render engine for NeRFs!

SCONE. (NeurIPS 2022 - Spotlight) Official code of SCONE: Surface Coverage Optimization in Unknown Environments by Volumetric Integration

pips. Particle Video Revisited

MaskingDepth. Python

blender-plots. Python library for making 3D plots with blender

opticalflow-autoflow. Jupyter Notebook

robustmvd. Repository for the Robust Multi-View Depth Benchmark

ml-neuman. Official repository of NeuMan: Neural Human Radiance Field from a Single Video (ECCV 2022)

MonoViT. Self-supervised monocular depth estimation with a vision transformer

kaolin-wisp. NVIDIA Kaolin Wisp is a PyTorch library powered by NVIDIA Kaolin Core to work with neural fields (including NeRFs, NGLOD, instant-ngp and VQAD).

OnDA. Source code for "Online Unsupervised Domain Adaptation for Semantic Segmentation in Ever-Changing Conditions", ECCV 2022. This is the code has been implemented to perform training and evaluation of UDA approaches in continuous scenarios. The library has been implemented in PyTorch 1.7.1. Some newer versions should work as well.

Shallow_DA. Official Repository for "Shallow Features Guide Unsupervised Domain Adaptation for Semantic Segmentation at Class Boundaries"

Self-supervised-Monocular-Trained-Depth-Estimation-using-Self-attention-and-Discrete-Disparity-Volum. Reproduction of the CVPR 2020 paper - Self-supervised monocular trained depth estimation using self-attention and discrete disparity volume

CER-MVS. Python

3DVideos2Stereo. Code to extract stereo frame pairs from 3D videos, as used in "Ranftl et. al., Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer, arXiv:1907.01341"

HashNeRF-pytorch. Pure PyTorch Implementation of NVIDIA paper on Instant Training of Neural Graphics primitives: https://nvlabs.github.io/instant-ngp/

Graft-PSMNet. Python

nice-slam. [CVPR'22] NICE-SLAM: Neural Implicit Scalable Encoding for SLAM

neural-disparity-refinement. Python

360OpticalFlow-TangentImages. The implementation of the BMVC 2021 paper 360 Optical Flow using Tangent Images.

CREStereo. Official MegEngine implementation of CREStereo(CVPR 2022 Oral).

TensoRF. [ECCV 2022] Tensorial Radiance Fields, a novel approach to model and reconstruct radiance fields

kubric. A data generation pipeline for creating semi-realistic synthetic multi-object videos with rich annotations such as instance segmentation masks, depth maps, and optical flow.

semantic_nerf. The implementation of "In-Place Scene Labelling and Understanding with Implicit Scene Representation" [ICCV 2021].

mipnerf. Python

svox2. Plenoxels: Radiance Fields without Neural Networks

EPCDepth. [ICCV 2021] Excavating the Potential Capacity of Self-Supervised Monocular Depth Estimation

implicit-slam. Dense SLAM with an Implicit Neural Representation

IterMVS. Official code of IterMVS (CVPR 2022)

arxiv-sanity-lite. arxiv-sanity lite: tag arxiv papers of interest get recommendations of similar papers in a nice UI using SVMs over tfidf feature vectors based on paper abstracts.

aioway. AI on the way. An auto deep learning pipe dream. An RDBMS approach to deep learning. Declarative, explainable, scalable, optimizable, easy to deploy, all that good stuff.

Esito. Esito ambition is to be your return type for suspending functions.

solvability. Federica Arrigoni, Andrea Fusiello, Elisa Ricci and Tomas Pajdla. Viewing Graph Solvability via Cycle Consistency. ICCV 2021

d4-dbst-old. Python

SeparableFlow. Separable Flow: Learning Motion Cost Volumes for Optical Flow Estimation

Q-Match. This is the code for the experiments in the ICCV publication 'Q-Match: Iterative Shape Matching using Quantum annealing' in http://gvv.mpi-inf.mpg.de/projects/QMATCH. You can run the code with D-Wave leap: https://www.dwavesys.com/take-leap. One can either run code with the leap IDE or locally by installing ocean (https://docs.ocean.dwavesys.com/en/stable/getting_started.html). To get to the Leap IDE one has to click on workspaces after logging in to Leap.

RAFT-Stereo. Python

HITNET-Stereo-Depth-estimation. Python scripts for performing stereo depth estimation using the HITNET Tensorflow model.

manydepth. [CVPR 2021] Self-supervised depth estimation from short sequences

mobilestereonet. Lightweight stereo matching network based on MobileNet blocks

PatchmatchNet. Official code of PatchmatchNet (CVPR 2021 Oral)

coex. Python

SynLiDAR. SynLiDAR: Synthetic LiDAR sequential point cloud dataset with point-wise annotations (AAAI2022)

torch_kitti. PyTorch utilities to handle the KITTI Vision Benchmark Suite

openpylivox. Python3 driver for Livox lidar sensors

MonoRec. Official implementation of the paper: MonoRec: Semi-Supervised Dense Reconstruction in Dynamic Environments from a Single Moving Camera (CVPR 2021)

supermario-dqn. Deep Reinforcement Learning Agent for Super Mario Bros

DICL-Flow. [NeurIPS 2020] Displacement-Invariant Matching Cost Learning for Accurate Optical Flow Estimation

MS-Nets. Matching space stereo networks - MSNet, with improved generalization properties

RAFT-3D. Python

SMD-Nets. SMD-Nets: Stereo Mixture Density Networks

DPT. Dense Prediction Transformers

uncertainty-toolbox. Uncertainty Toolbox: a Python toolbox for predictive uncertainty quantification, calibration, metrics, and visualization

awesome-NeRF. A curated list of awesome neural radiance fields papers

compass. Repository containing the code of "Learning to Orient Surfaces by Self-supervised Spherical CNNs".

LAF. LAF-Net, CVPR'19 (oral)

rectified-features. [ECCV 2020] Single image depth prediction allows us to rectify planar surfaces in images and extract view-invariant local features for better feature matching

footprints. [CVPR 2020] Estimation of the visible and hidden traversable space from a single color image

stereo-from-mono. [ECCV 2020] Learning stereo from single images using monocular depth estimation networks

DenseMatchingBenchmark. Dense Matching Benchmark

AcfNet. [AAAI2020] Adaptive Unimodal Cost Volume Filtering for Deep Stereo Matching

Reversing. Code for "Reversing the cycle: self-supervised deep stereo through enhanced monocular distillation"

DepthComplete. Pytorch implementation of depth completion architectures (eg. SparseConv, Sparse-to-Dense)

demo_live. You can try the demo here:

omeganet. Distilled Semantics for Comprehensive Scene Understanding from Videos [CVPR 2020]

netdef-docker. DispNet3, FlowNet3, FlowNetH, SceneFlowNet -- in Docker

depth-hints. [ICCV 2019] Depth Hints are complementary depth suggestions which improve monocular depth estimation algorithms trained from stereo pairs

ICCV_19. Federica Arrigoni and Tomas Pajdla. Robust Motion Segmentation from Pairwise Matches. ICCV 2019

DSMNet. Domain-invariant Stereo Matching Networks

DWARF-Tensorflow. TensorFlow implementation of "Learning end-to-end scene flow by distilling single tasks knowledge"

flowattack. Attacking Optical Flow (ICCV 2019)

briefmatch. BriefMatch real-time GPU optical flow

optical-flow-filter. A real time optical flow algorithm implemented on GPU

ATDT. Implementation of "Learning Across Tasks and Domains" ICCV 2019

d2-net. D2-Net: A Trainable CNN for Joint Description and Detection of Local Features

mobilePydnet. Pydnet on mobile devices

LGC-Tensorflow. We propose to exploit nearby and farther clues available from image and disparity domains to obtain a more accurate confidence estimation. While local information is very effective for detecting high frequency patterns, it lacks insights from farther regions in the scene. On the other hand, enlarging the receptive field allows to include clues from farther regions but produces smoother uncertainty estimation, not particularly accurate when dealing with high frequency patterns. For these reasons, we propose a multi-stage cascaded network to combine the best of the two worlds.

Learning2AdaptForStereo. Code for: "Learning To Adapt For Stereo" accepted at CVPR2019

monoResMatch-Tensorflow. Tensorflow implementation of monocular Residual Matching (monoResMatch) network.

Semantic-Mono-Depth. Geometry meets semantics for semi-supervised monocular depth estimation - ACCV 2018

Real-time-self-adaptive-deep-stereo. Code for "Real-time self-adaptive deep stereo" - CVPR 2019 (ORAL)

pydnet. Repository for pydnet, IROS 2018

lautaro. Kotlin

Unsupervised-Confidence-Measures. This strategy provides labels for training confidence measures based on machine-learning technique without ground-truth labels (BMVC 2017)

Unsupervised-Adaptation-for-Deep-Stereo. Code for "Unsupervised Adaptation for Deep Stereo" - ICCV17

CrossScaleStereo. Cross-Scale Cost Aggregation for Stereo Matching (CVPR 2014)

models. Models and examples built with TensorFlow

My-Photo-Diary. Demo application using cordova and ionic.