A REST API for Structure from Motion (SfM) reconstruction using Hierarchical Localization (hloc) and COLMAP.

A library to load scenes for "gausplat"

The implemention of the NeRF architecture from scratch that can be trained using synthetic data for 3D model rendering.

Small script for easier image adding in colmap

Python CLI/Library to process videos, images, colmap scenes to select the best frames of a 3D capture

This repository contains Python code for image matching, feature detection, and 3D reconstruction using COLMAP, a popular computer vision toolkit.

Cinema 4D Python importer for COLMAP/GLOMAP camera tracking. Builds Redshift-compatible animated cameras, imports sparse point clouds, and configures scene settings for seamless photogrammetry workflows—an alternative to Blender’s photogrammetry importer, adapted for C4D.

Toolkit for a collaborative and shardable repository of the 2d & 3d likeness of objects.

Website for Depth Supervised Neural Scene Flow

Senior project with Austin Theriault and Ryan Hudson

The aim of my project was to develop a user-friendly method for generating realistic 3D models and displaying them on smartphones using augmented reality. The only input required from the user is a short video of the desired object, which can be uploaded to the app. This project uses a modified version of Nvidia's Instant-NeRF.

A repo for my Graduate Computer Vision I semester project

This repository contains the code developed for my MSc project titled "Exploring Photogrammetry for High-Fidelity 3D Model Generation: A Power Grid Inspection Study", as well as some of the results. The thesis was written in collaboration with SINTEF, and the degree issued by DTU. The full thesis can be downloaded from the below link.

Nix packaging of XiaoBaiiiiii/colmap-pcd. Work in progress.

A Python-based photogrammetry pipeline for 3D reconstruction using COLMAP and Open3D. Features include feature extraction, matching, sparse/dense reconstruction, and texture mapping.

A Python-based Structure from Motion implementation that reconstructs 3D scenes from image sequences. Key features: - SIFT feature detection and matching - Graph-based image selection - Incremental reconstruction - Bundle adjustment optimization - COLMAP/PLY export support Perfect for computer vision researchers and 3D reconstruction enthusiasts.

a streamlined pipeline for converting videos into 3D Gaussian Splatting models

Provides a function to delete image ids from points3D.txt file for COLMAP

Reconstruction 3D image with colmap