lolleko/mesh-data-synthesizer
Uses Unreal Engine & Cesium to generate large synthetic dataset from 3D meshes. Enables machine learning tasks like Visual Place Recognition read more in our paper on this: https://meshvpr.github.io
This tool helps researchers and developers create large, synthetic datasets from detailed 3D mesh models of real-world locations. You provide a 3D model (like a city or landscape) and the tool generates thousands of photorealistic images from various viewpoints within that model. This data is perfect for training machine learning systems to recognize specific places from images, enabling applications like robot navigation or augmented reality.
No commits in the last 6 months.
Use this if you need to generate a vast number of diverse, high-quality images from 3D geospatial models to train visual AI for tasks like recognizing locations.
Not ideal if you need to generate data types other than RGB images, require extremely high stability for production use, or want to easily integrate custom data generators.
Stars
35
Forks
3
Language
C++
License
—
Category
Last pushed
Jun 20, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/lolleko/mesh-data-synthesizer"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
MediaTek-NeuroPilot/mai21-learned-smartphone-isp
The official codebase for the Learned Smartphone ISP Challenge in MAI @ CVPR 2021
ashishpatel26/365-Days-Computer-Vision-Learning-Linkedin-Post
365 Days Computer Vision Learning Linkedin Post
extreme-assistant/ICCV2023-Paper-Code-Interpretation
ICCV2021/2019/2017 论文/代码/解读/直播合集，极市团队整理
amusi/daily-paper-computer-vision
记录每天整理的计算机视觉/深度学习/机器学习相关方向的论文
extreme-assistant/survey-computer-vision-2020
2020-2021年计算机视觉综述论文分方向整理