Tagged articles
point cloud
11 articles · Page 1 of 1
This tutorial surveys 3D vision-language models, covering geometric representations (point clouds, meshes, Gaussian splatting), CLIP-style cross-modal alignment, taxonomy of embedding-based vs. generative architectures, and emerging applications in recognition, scene understanding, 3D generation, and embodied AI, highlighting the distinction between similarity matching and language generation capabilities.
3D Gaussian splatting3D vision-language modelsCLIP
0 likes · 29 min read

This article reviews the MPCT framework—a multiscale point‑cloud transformer built on a residual network that leverages permutation‑invariant self‑attention, point‑enhancement, and hierarchical feature aggregation to achieve state‑of‑the‑art results on ModelNet40 and ScanObjectNN datasets.
3D classificationSelf-Attentionmultiscale
0 likes · 14 min read

SIF3D is a scene‑aware 3D human motion forecasting framework that fuses observed motion, 3D point‑cloud scenes, and gaze through novel ternary intention‑aware and semantic‑coherence‑aware attention mechanisms, encoding with PointNet++ and Transformers, and decoding with a graph‑convolutional network, achieving state‑of‑the‑art results on GIMO and GTA‑1M benchmarks.
3D Scene UnderstandingCVPR2024Computer Vision
0 likes · 15 min read

This article presents a Python project that reads point‑cloud data from CSV files, applies DBSCAN and KMeans clustering, extracts cluster features, and uses the Hungarian algorithm to match clusters across frames for tracking, complete with full source code and result visualization.
DBSCANHungarian algorithmKMeans
0 likes · 13 min read

IPNet introduces a hybrid Image‑Point Cloud architecture that first extracts 2D visual features with a CNN, projects them into 3D point‑cloud space, and iteratively refines hand pose using a sparse‑anchor “aggregate‑interact‑propagate” scheme, achieving state‑of‑the‑art results on challenging hand‑object datasets.
2D-3D fusionAISemGCN
0 likes · 6 min read

This article introduces eight Python libraries—Open3D, Trimesh, PyVista, Vedo, Pyrender, PlotOptiX, Polyscope, and Simple‑3dviz—detailing their installation, core APIs, and step‑by‑step code examples for visualizing meshes, point clouds, and multidimensional data.
3D visualizationPythonmesh
0 likes · 41 min read

This tutorial introduces the Open3D library, shows how to install it, convert depth maps to point clouds, visualize facial point clouds, perform point‑cloud segmentation and clustering, and apply surface reconstruction methods such as Poisson, alpha shapes, and ball‑pivoting using Python code.
3D processingPythonopen3d
0 likes · 12 min read

The Eagle.gl Web 3D engine, built atop Three.js, provides a unified 2D/3D GIS platform that efficiently loads, renders, and edits massive point‑cloud, vector, and model map data with real‑time LOD, GPU picking, DEM terrain integration, and configurable styling, supporting AMap’s HD‑map production and future autonomous‑driving simulation.
3D renderingGISThree.js
0 likes · 12 min read

The article describes how high‑frequency LiDAR scanners on precision mapping vehicles suffer from Multi‑Time‑Around (MTA) errors—mis‑assigning distant returns to near ranges—and explains both internal laser strategies (continuity assumption and variable‑period emission) and a four‑step neighborhood‑weighting algorithm that reliably corrects these artifacts, restoring accurate point‑clouds for automated map generation.
LiDARMTAdata processing
0 likes · 12 min read

This article explains the principles, components, generation process, and challenges of high‑precision topological and point‑cloud maps, and describes satellite‑based, map‑based, and fused high‑precision localization methods that underpin perception, prediction, planning, and control in autonomous driving systems.
Autonomous DrivingSLAMhigh-precision mapping
0 likes · 9 min read

With 5G accelerating video traffic, immersive media relies on advanced 3D representation, efficient point‑cloud and video‑based compression, standardized containers, and selective streaming to enable realistic VR/AR experiences across gaming, autonomous driving, and broadcasting, while Tencent’s lab drives standards and future XR applications.
3D5GVR
0 likes · 12 min read
