DataGym.aihat einen Beitrag geschrieben.20. Oktober 2020
Datagym AI Series Part 5: When human agents deal with uncertainty in decision making they usually have many different options to gain more information in order to increase their certainty. Today's video gives a quick overview of how this handling of uncertainty translates into AI solutions.
https://www.datagym.ai/ https://www.youtube.com/watch?v=MWlQUtI2NOg
DataGym.aihat einen Beitrag geschrieben.31. August 2020
Today we show you what a PR-Curve is.
Precision is the fraction of retrieved instances that are relevant, while recall is the fraction of retrieved instances to all relevant instances. But what does that mean? Our machine learning expert explains as simple as possible, what a PR-Curve is.
Have fun watching it!
https://www.youtube.com/watch?v=QCYaOyJ1rcc
DataGym.aihat einen Beitrag geschrieben.17. August 2020
Labeling images can be time intensive and expensive, especially for instance segmentation. DATAGYM.AI can speed up the label task significantly by pre-labeling images with polygons. Correcting polygons is easier and faster than labeling them from scratch after all.
https://www.youtube.com/watch?v=huRG50kttxo&feature=youtu.be
DataGym.aihat einen Beitrag geschrieben.28. Juli 2020
AI Series Part 2 - Perspective in data
Our second part of the AI series is ready! Training Neural Networks requires data, a lot of it. But why? Today we show how different view angles on 3D objects contribute to this need for data.
https://youtu.be/LgJSL2pmj0E
DataGym.aihat einen Beitrag geschrieben.20. Juli 2020
New AI Series!
In our new learning AI series we will be giving a broad overview on Machine Learning projects. Today we have a short introduction to iterative loops within the AI workflow.
Have fun watching it!
https://youtu.be/nON366hCeyI
DataGym.aihat einen Beitrag geschrieben.16. Juli 2020
You always wanted to know what DATAGYM, AI is?
Label up to 10x faster!
DATAGYM enables data scientists and machine learning experts to label images up to 10x faster. AI-assisted annotation tools reduce manual labeling effort, give you more time to finetune ML models and speed up your go to market of new products.
www.datagym.ai https://youtu.be/0n9CFE3etj8
DataGym.aihat einen Beitrag geschrieben.14. Juli 2020
DATAGYM.AI is now supporting AWS S3 as file storage. Amazon S3 is a simple, reliable and cost-effective object storage service that provides nearly limitless, highly reliable cloud storage capacity. Many customers store and manage all their data on S3. DATAGYM.AI now allows you to connect your S3 buckets to datasets within your DATAGYM account. https://www.datagym.ai/ https://youtu.be/1kZXOtAC1V4
DataGym.aihat einen Beitrag geschrieben.25. Juni 2020
New Tutorial:
Labeling images can be time intensive and expensive. By leveraging object detection networks for pre labeling of your data, DATAGYM can speed up the label task significantly. Correcting labels is easier and faster than labeling from scratch after all. Have a look how our Team Pro users can use DATAGYM to pre label with up to 80 distinct object classes.
https://youtu.be/9XVMFzd8GuA
DataGym.aihat einen Beitrag geschrieben.17. Juni 2020
New Tutorial: "How to correct AI-assisted polygon!"
In the previous posting we have showed you how you can use our AI assisted labeling tool. But what happens when this network makes a small mistake? Today we’ll show you how you can refine the polygon by sending additional points as input so that the network can correct itself. Try it on datagym.ai.
https://youtu.be/Fvt7QKy9EZc
