We added a new product today, and it's available for free download. RevEyes ML is a Windows application that can be used to edit and retrain computer vision ML (machine learning) models.

Computer vision ML models are frequently trained to recognize particular classes of objects. For example, models trained on the "imagenet" image set can recognize any of the 1000 classes of objects (such as cars, various breeds of dogs, etc.) that are part of imagenet. When you need an ML model that performs the same function as an existing model, but you need to identify different classes of objects (for example, your mom and dad, correctly vs incorrectly assembled products, etc.,) a process called Transfer Learning can be used to update the model to meet your specific needs.

Unfortunately, the process of retraining a model can be fairly complicated -- far more complicated than we believe it should be. The typical process involves installing Python, installing a variety of Python libraries (ensuring that you have the correct versions of each so that they can work together,) writing a bunch of Python code to edit the structure of the model, and then writing even more code to retrain the model using images you supply of the object classes you want it to recognize.

We created RevEyes ML to make this process easier.

RevEyes ML is a Windows program that can load, edit, and retrain computer vision models without requiring the installation of Python or the writing of any code.

Our initial goal was to provide a tool for users of OptimEyes Developer to train models for use with its ML Tool. However, we decided that rather than bundling RevEyes ML with OptimEyes Developer as a commercial product, we'd release it for free to anyone who wants to retrain ML models on Windows without having to deal with the complexity of managing a Python environment and writing code.

If you like RevEyes ML -- if you enjoy skipping all the code writing -- check out OptimEyes Developer. It does for computer vision solution development what RevEyes ML does for retraining ML models: it provides a Windows environment where you can develop and deploy complete machine vision solutions (both ML-based and using traditional computer vision techniques) without ever having to write a line of code.