Teach machines to see the world in three dimensions. These guides cover the algorithms that turn raw pixels and points into geometric understanding, from feature matching to pose estimation.
Label a 3D point cloud with any word you type. This complete guide walks the open-vocabulary 3D segmentation stack in Python: pretrained SAM cuts the scene, CLIP scores each region, and a scene graph lets an LLM answer questions about the space.
Open-Vocabulary 3D Segmentation Complete Guide Read More »
A complete, practical 101 on 3D point cloud segmentation in Python: what it is, the method families, real Open3D code for RANSAC and DBSCAN, and the parameters that decide your results.
3D Point Cloud Segmentation Complete Guide Read More »
A complete Python pipeline for 3D point cloud labeling from phone photos using interactive painting, pinhole projection, and KD-tree fusion — no GPU required.
3D Point Cloud Labeling from Photos with Python (and why AI can’t design it for you) Read More »
By Dr. Florent Poux, PhD in 3D geospatial sciences | Author, 3D Data Science with Python (O’Reilly) | Published 30 June 2025 A 3D scene graph encodes what a point cloud cannot: how objects relate to each other, that a lamp sits on a table or a bookshelf stands against a wall. Building one in […]
3D Scene Graphs for Spatial AI with NetworkX and OpenUSD Read More »
