Creating machine learning models using Lobe
2 min read
Create your own machine learning model using no-code.
Yes, you heard that right. Lobe enables you to create machine learning models using no-code that automatically train themselves and improve with each new dataset. Lobe aims to make machine learning easy and accessible for everyone.
What is Lobe?
It is easy to use, download the application and start labeling your images. The models will be trained on your own computer without the need to upload data to the cloud. Lastly you can export your model to multiple platforms.
Currently it is in beta with only image classification being available, but there are more templates coming for object detection and data classification.
Create machine learning models
Label using images, camera or imported dataset
Use through image, camera or exporting
Local training, no need to upload to cloud
Start training your model with as few as 5 images and let Lobe do its magic. Lobe works in the background improving the model, and with each feedback from the user, it gets higher accuracy. So, the bigger dataset you use to train your models, the better accuracy you will have.
Currently it is in beta and free to use.
What did I create?
I created 2 labels, one being: drinking glass(not feeling creative today) and the other: not drinking glass. The idea was to have a model that differentiates drinking glasses to other objects. Everything from labelling of the images using my camera, to the training of the model and testing was done in only 10 minutes. It is incredible how you can have a working model in such a short time!
Here is a preview of the model in action.
My experience with Lobe
Ease of use 5/5
Community info 4/5
Code required: none
Do you have any experience creating machine learning models. If yes, let me know what you think of Lobe?