AI Learning Environment

Morpho’s AI learning environment implements a variety of models, training algorithms, and typical processing functions required for training Deep Neural Networks for Machine Learning tasks such as classification, detection, segmentation, etc. It enables AI developers to perform a variety of training tasks just by without new implementations by changing options, without re-implementing these algorithms.

Use Case

Function or Purpose

Features

You can perform training in Tensorflow, without having to implement specific training algorithms.
It can be used for various tasks such as classification, detection, segmentation.
A large number of high-speed augmentation processes are implemented, enabling full-scale Training.

Product Details

When you train DNN-based AIs, you usually use frameworks such as Tensorflow.
However, in order to perform actual Training using the framework, you still have to implement a number of things such as; preprocessing images and annotations into a format that the framework can handle, performing data augmentation before/during training, preparing models for neural networks, and preparing loss functions and optimizers for learning.
In addition, when you change the learned content (tasks and models), you generally need to re-implement them.
Morpho’s AI learning environment implements a variety of models, such as the typical processes required for training tasks such as classification, detection, segmentation, allowing you to perform a variety of training by simply changing a few options, without new implementations.