Create Object Detection and Segmentation Neural Networks without Code

MakeML trains ML models
MakeML is built to make training process easy to setup. It is designed to handle data sets, training configurations, markup and training process. All in one place.

ML project configuration

Select Object Detection or Semantic Segmentation Neural Network type and create your training project in minutes. MakeML supports Tensorflow and Turicreate frameworks with CoreML and TFlite models available as a result.

MakeML project configurations help you to run ML model training without spending a lot of time trying to setup python dependencies. You can also change all necessary training parameters, such as batch size, learning rate, and the number of iterations.

Data sets markup

Harness the full power of the MakeML Markup Tool to label your dataset. No need to search for dataset converters anymore - MakeML prepares your converts your dataset to the right format right before training.

MakeML was designed to prepare your data sets for training. Import existing dataset if you have one and MakeML will automatically convert it to needed format, or import images and label them inside MakeML. You can also export your dataset if you need to share it or use somewhere else.

Training Model

Instead of setting-up cloud instances and creating your training pipeline, press run button to train your model on powerful MakeML GPU instance in the cloud.

MakeML gives you possibility to see all output during training. You can even use MakeML iOS app to track your trainings while you're not at your desk.

ML experiment management

Features

Select Object Detection or Semantic Segmentation Neural Network type and create your training project in minutes. MakeML supports Tensorflow and Turicreate frameworks with CoreML and TFlite models available as a result.

Harness the full power of MakeML Markup Tool to label your dataset. No need to search for dataset converters anymore - MakeML prepares your converts your dataset to right format right before training.

Instead of setting-up cloud instances and creating your training pipeline, press run button to train your model on powerful MakeML GPU instance in the cloud.

Quick guide to Machine Learning on Mobile

Apps that use Object Detection or Segmentation

MSQRD

MSQRD uses Object Detection to detect keypoints on your face and then apply 3D masks to your face. This technology also used in Snapchat masks.

Homecourt

Apple Design Awards winner app called Homecourt uses Object Detection to determine ball and player position, player posture, and basket position to create the best shot tracking basketball experience ever.

Vivino

Vivino uses Object Detection to allow you to take a photo of any wine label at a store and instantly see detailed information about the wine and all available purchasing options.

MSQRD

MSQRD uses Object Detection to detect keypoints on your face and then apply 3D masks to your face. This technology also used in Snapchat masks.

Homecourt

Apple Design Awards winner app called Homecourt uses Object Detection to determine ball and player position, player posture, and basket position to create best shot tracking basketball experience ever.

Vivino

Vivino uses Object Detection to allow you to take a photo of any wine label at a store and instantly see detailed information about the wine and all available purchasing options

Have a question - book a call

How it works
Do you want to know how to train Object Detection CoreML model in 3 minutes? Watch the video!

Subscribe for updates. We won't spam you.

What our Clients Say About Us
On collecting and processing data
"I like the image labeling feature and the speed with which you improve it."
- Fabien
“If you add an automated labeling tool to your macOS app, I will buy 2 yearly subscriptions!”
- Alex
On setup training pipeline
"The app itself, brilliant. Takes away a lot of complexity."
- Mark
"Your app made my life so much easier!"
- Kevin
On integration of a model
"I really want to learn more about ML in mobile apps, and I think you will have no issue finding interested customers!"
- @harrissonkolts, Reddit
"I know plenty would be curious about learning to use Core ML 3 and being walked through."
- Tanner

Quick guide to Machine Learning on Mobile