#3d-reconstruction

From Photos to Playable Mesh: A Technical Primer on Modern 3D Scene Reconstruction

How structure-from-motion and mesh-based Gaussian splatting turn an ordinary folder of photos into a real-time, exportable 3D scene

VGGT and D4RT

The best papers of CVPR in 2025 and 2026 are VGGT: Visual Geometry Grounded Transformer and Efficiently Reconstructing Dynamic Scenes One…

Turn Photos Into Measurable 3D: Meet OpenReco

An open-source photogrammetry and 3D reconstruction platform that takes a folder of drone photos and hands you a georeferenced, measurable…

3D Reconstruction Used to Require Calibrated Cameras.

The original VGGT was powerful but memory-hungry. VGGT-Ω replaces per-frame processing with compact registers that share scene information…

3D Reconstruction from Scratch, Part 4: Structure-from-Motion, Bundle Adjustment & Dense…

In the first three parts of this series, we slowly built the foundation of a 3D reconstruction pipeline from scratch. We started with…

3D Reconstruction from Scratch, Part 3: Camera Pose Estimation & Triangulation

In the last two parts of this series, we built the foundation for a 3D reconstruction pipeline from scratch. We started with feature…

Some Basics on 3D Reconstruction

AI ပိုင်းလိုက်တဲ့သူတွေ၊ data သမားတွေဆိုရင် ကိုယ် မသိထားတဲ့ data file types ဆိုတာ တစ်ချိန်ချိန်မှာ မလွဲမသွေကြုံရမှာပါ။ အဲ့ထဲမှာ 3D model…

I Built a Tool That Lets You Step Back Inside a Room Using Just Your Phone

Imagine this.

3D Reconstruction from Scratch, Part 2: Feature Matching and the Search for Correspondence

In Part 1, we focused on feature extraction, the step where a machine learns where to look in an image. We saw how algorithms like SIFT…

3D Reconstruction from Scratch, Part 1: Feature Extraction and the Search for Distinctive Points

Imagine taking a few photos of the same building from different angles. To us, it is easy to notice the same window corners, edges, and…

[Google DeepMind / UC Berkeley] — Hybrid Memory를 통한 장기 컨텍스트 기하학적 재구성 모델 ‘LoGeR’

배경 및 개요

My Journey into 3D reconstruction of people: Part 3 (Photogrammetry)

This is part of my blog series that covers my attempts at reconstructing a 3D human from images or videos as part of my larger Virtual…

My journey into 3D reconstruction of people: Part 2 (MapAnything)

This is part of my blog series that covers my attempts at reconstructing a 3D human from images or videos as part of my larger Virtual…

My journey into 3D reconstruction of people: Part 1 (from single image using GLPN transformers)

Whenever I do online shopping in fashion stores, I have wondered if it was possible to have a virtual trial room to aid in our dress…

Application of Plucker coordinates for computer vision/graphics

In this short article, I will be discussing about the application of Plucker coordinates in the domain of computer vision and graphics…

What Happens If You Initialize 3D Gaussian Splatting From a Very Sparse Point Cloud?

As you might know, most 3D Gaussian Splatting (3DGS) pipelines are initialized from a point cloud-typically produced by COLMAP/SfM. When…

Extracting Depth From Any Pretrained 3DGS (No Retraining Needed)

Extract depth maps from any pretrained 3D Gaussian Splatting without retraining, depth loss, or special…

From Images to 3D Worlds

Spatial Artificial Intelligence: The New Frontier

Rebuilding Maps from Pixels: The Power of Image-Driven Reconstruction

AI in Archaeology: What Surprised Me the Most

I’ve followed the development of AI in business, healthcare, logistics — the usual areas. But recently I stumbled upon a field we rarely…

3D Gaussian Splatting Tutorial from Scratch in 100 lines of PyTorch code (no CUDA, no C++)

Learn how 3D Gaussian Splatting works — implemented entirely in PyTorch, simple and fast to experiment with

SIFT: Feature Detection That Revolutionized Image Matching

Imagine a world where computers can effortlessly identify objects in images, regardless of changes in scale, rotation, or even lighting…

VGGT: How Feed-Forward 3D Perception is Redefining Scene Reconstruction and Multi-View Geometry

In June 2025, VGGT (Visual Geometry Grounded Transformer) claimed the Best Paper Award at the CVPR (Computer Vision and Pattern…

From Images to 3D: Building Im3D — A Depth Estimation Based Point Cloud and Mesh Generator

🚀 Introduction

Now in 3D: Flying the LEGO Plane with Yaw, Pitch, and Roll: Building Intuition for Nerf : Part4

So far, we’ve been watching Sally and Joe spin objects around in 2D — drawing circles, rotating arrows, and arguing about what’s clockwise…

LiDAR Meets Vision — Comprehensive Guide To 3D Reconstruction

3D reconstruction in computer vision facilitates the transformation of 2D visual inputs or point clouds into detailed three-dimensional…

SF3D: Elevating Single-Image 3D Reconstruction Beyond TripoSR

The ability to generate high-quality 3D models from a single image has immense potential across various domains, from gaming and virtual…

23. 뉴럴 렌더링 기반 VPS 공간맵 생성 기술

안녕하세요, 저는 MAXST에서 AI와 Neural Rendering 연구를 담당하고 있는 김상록입니다.

23. Neural Rendering-based VPS Spatial Map Generation

Hello, this is Shawn, researching AI and Neural Rendering at MAXST.

19. MAXST VPS 맵 구축 파이프라인 소개 (1/2)

안녕하세요, MAXST에서 VPS 를 연구하고 있는 이형민입니다.

19. Introducing the MAXST VPS Map Building Pipeline (1/2)

Hello, this is Paul, researching VPS at MAXST.

Instant NeRF on Google Compute Engine via Chrome Remote Desktop

Rendering a 3D NERF Toy Gun with Neural Radiance Fields (NeRF) on a Google Cloud VM