Computer Vision and Geometry Lab

LightGlue

LightGlue: Local Feature Matching at Light Speed (ICCV 2023)

Hierarchical-Localization

Visual localization made easy with hloc

nice-slam

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

pixel-perfect-sfm

Pixel-Perfect Structure-from-Motion with Featuremetric Refinement (ICCV 2021, Best Student Paper Award)

limap

A toolbox for mapping and localization with line features.

glue-factory

Training library for local feature detection and matching

SOLD2

Joint deep network for feature line detection and description

GlueStick

Joint Deep Matcher for Points and Lines 🖼️💥🖼️ (ICCV 2023)

DeepLSD

Implementation of the paper "DeepLSD: Line Segment Detection and Refinement with Deep Image Gradients"

visloc-iccv2021

ETH-Microsoft dataset for the ICCV 2021 visual localization challenge

pyceres

Factor graphs with Ceres in Python

nicer-slam

[3DV'24 Best Paper Honorable Mention] NICER-SLAM: Neural Implicit Scene Encoding for RGB SLAM

VP-Estimation-with-Prior-Gravity

Vanishing Point Estimation in Uncalibrated Images with Prior Gravity Direction (ICCV 2023)

Mask3D

Mask3D predicts accurate 3D semantic instances achieving state-of-the-art on ScanNet, ScanNet200, S3DIS and STPLS3D.

LabelMaker

raybender

Fast CPU rendering in Python using the Intel® Embree backend

pcdmeshing

Point cloud meshing with CGAL

hololens_ros

mav_active_3d_planning

Modular framework for online informative path planning.

sdfstudio

A Unified Framework for Surface Reconstruction

spot_pose_estimation

Grounded-Segment-Anything

Packaged version of Grounded-SAM to be used in LabelMaker

open3d-manylinux2014

This is a build code for building open3d wheels on manylinux 2014 docker, so that ETHz's Euler (CentOS 7) can run it.

spot_ros2

ROS2 driver package for boston dynamic's spot

CVG-Thesis-Template

maplab

cvg development fork of maplab (to be able to create pull requests)

maplab_msgs

Custom ROS messages for maplab

maplab_tools

OnePose_Plus_Plus_Spot

Code for "OnePose++: Keypoint-Free One-Shot Object Pose Estimation without CAD Models" NeurIPS 2022

recognize-anything

Packaged version of Recognize Anything to be used in LabelMaker