Synthetic Object Recognition Dataset for Industries.
News
-
A NEW book is released discussing the impact of digitalization and digital transformation technologies on the Industry 4.0 and smart factories, how the factory of tomorrow can be designed, built, and run virtually as a digital twin likeness of its real-world counterpart, before the physical structure is actually erected. For more info: [Link 1] [Link 2]
-
Experience the power of innovation with GREEN Physics AI and Metadata, revolutionizing the way we monitor, analyze, and optimize assets for maximum performance and efficiency!
-
c’t 20/2022
-
BMW PRESSCLUB GLOBAL
-
NVIDIA GTC 2022
LATEST NEWS:
Together with Vertex AI, SORDI.ai Revolutionizes Industrial Processes
SORDI.ai is collaborating with Google Cloud to convert real-world assets into precise digital counterparts, so-called “digital twins” - to digitize planning and optimize supply chain with Generative AI. For more information, check the official press release!
"The results of our development collaboration (with Google Cloud) in the context of SORDI.ai define new standards and significantly accelerate the process of creating synthetic industrial data."
BMW Group
"[…] Simulations of manufacturing processes can be easily streamlined with a growing number of 3D models."
Google Cloud
The largest and most comprehensive synthetic dataset
Published by BMW Group, SORDI.ai helps developers and researchers to streamline and accelerate the training of artificial intelligence in production. Together with Microsoft, NVIDIA, and idealworks, SORDI.ai has been made available as open-source, in an effort to build the world’s largest reference dataset for artificial intelligence in the field of manufacturing.
120+ unique object classes
SORDI.ai is a synthesised data set containing over 100 object classes, and over 1.2 Million photorealistic images and their modalities. SORDI.ai is designed to suit general image processing tasks including classification, object detection or segmentation.
-
Over 800,000 synthetically generated images
-
Photorealistic renders and GAN generated images
-
Segmented and labelled images with bounding boxes
Over 80 versatile object classes
-
Logistics
Containers, silage, storage boxes, disposable and reusable handling materials, packaging, boxes.
-
Transportation
Manual and powered vehicles, pushbikes, scooters, production equipment.
-
Office Settings
Office furniture, boards, displays, accessories.
-
Signage
Emergency signs, information signs, pictographs, text signs.
Compare SORDI.ai image outputs
Click and slide to compare two different outputs:
SORDI.ai for Visual Inspection
SORDI.ai can be used to detect missing or wrong (colored) stitches in leather products and automate the visual inspection for valid leather and stitching color combinations. To showcase that we added a large number of images of leather products with different stitching patterns, textures, and colors, making it possible to train machine learning algorithms to identify missing or incorrect stitches in leather products automatically.
As a result, SORDI.ai enables manufacturers to ensure that their products are of high quality and meet the required standards.
Digital first approach.
The SORDI.ai dataset is made up of synthetic images generated using NVIDIA Omniverse. Through leveraging USD and MDL workflows, as well as connecting DCC tools from BMW Groups workflow, SORDI.ai is constantly expanding to include new models and classes.
Partners
Contributors
FAQs
What is the best way for us to get access to this dataset?
SORDI.ai is going to be released later this year. Please follow us on our GitHub account or fill out the contact form below, to keep up to date.
How is SORDI.ai better than real images?
SORDI.ai provides setting/quality coherent images from different viewports, angles, and complex human access areas! A thing that cannot be easily achieved in a real-life.
What CUDA and Tensorflow versions does our Github tools support?
They support CUDA 11 and TensorFlow 2.
What labeling format is SORDI.ai using in its annotations?
BMW has created its own advanced labeling format! All details can be found on our Github account: github.com/BMW-InnovationLab
Are there tools to use SORDI.ai for training?
We provide famous complete open-source tools for an AI training pipeline! From image labeling to model training and then evaluation! Feel free to check them on our GitHub account: github.com/BMW-InnovationLab
Are you using SORDI.ai in BMW to train your models?
Of course!
Get in touch.
If you would like to find out more about this project, request a class or feature, or contribute to our ongoing research — please connect with us.