ML.NET Model Builder Update Boosts Image Classification -- Visual Studio Magazine

ML.NET Model Builder Update Boosts Image Classification

By David Ramel
11/13/2019

Microsoft announced an update to the Model Builder component of its ML.NET machine learning framework, boosting image classification and adding "try your model" functionality for predictions with sample input.

Model Builder is the Visual Studio UI tooling that .NET developers use to work with the ML.NET cross-platform machine learning framework, which hit version 1.0 in May. "Developers with no ML expertise can use this simple visual interface to connect to their data stored in files or SQL Server, train the model and generate code for model training and consumption," Microsoft says. The ML.NET framework also works with a CLI (command-line interface) for use outside of the Visual Studio IDE, and AutoML, Automated Machine Learning, the Microsoft Research project used to automate the process of building custom models.

In an update post published yesterday (Nov. 12), program manager Bri Achtman announced the two new primary features:

Image classification scenario -- locally train image classification models with your own images: "We showed off this feature in .NET Conf to classify the weather in images as either sunny, cloudy, or rainy, and now you can locally train image classification models in Model Builder with your own images!" Note that image classification doesn't work with the ML.NET CLI.
Try your model -- make predictions on sample input data right in the UI: "After training a model in Model Builder, you can use the model to make predictions on sample input right in the UI for both text and image scenarios. For instance, for the dog vs. cat image classification example, you could input an image and see the results in the Evaluate step of Model Builder."

In addition to image classification, ML.NET is used for sentiment analysis, customer segmentation, sales spike detection, product recommendation, object detection and several other machine learning scenarios. Its Model Builder development is carried out on GitHub, and Model Builder is available on the Visual Studio Marketplace -- or it can be installed via the Visual Studio Extensions menu.

Model Builder (still in preview) has been installed more than 54,000 times, according to its marketplace entry, earning a 4.6 (0-5 scale) from 19 developers who reviewed it.

About the Author

David Ramel is an editor and writer at Converge 360.

Printable Format

comments powered by Disqus

Featured

Hands On: New VS Code Insiders Build Creates Web Page from Image in Seconds

New Vision support with GitHub Copilot in the latest Visual Studio Code Insiders build takes a user-supplied mockup image and creates a web page from it in seconds, handling all the HTML and CSS.
Naive Bayes Regression Using C#

Dr. James McCaffrey from Microsoft Research presents a complete end-to-end demonstration of the naive Bayes regression technique, where the goal is to predict a single numeric value. Compared to other machine learning regression techniques, naive Bayes regression is usually less accurate, but is simple, easy to implement and customize, works on both large and small datasets, is highly interpretable, and doesn't require tuning any hyperparameters.
VS Code Copilot Previews New GPT-4o AI Code Completion Model

The 4o upgrade includes additional training on more than 275,000 high-quality public repositories in over 30 popular programming languages, said Microsoft-owned GitHub, which created the original "AI pair programmer" years ago.
Microsoft's Rust Embrace Continues with Azure SDK Beta

"Rust's strong type system and ownership model help prevent common programming errors such as null pointer dereferencing and buffer overflows, leading to more secure and stable code."
Xcode IDE from Microsoft Archrival Apple Gets Copilot AI

Just after expanding the reach of its Copilot AI coding assistant to the open-source Eclipse IDE, Microsoft showcased how it's going even further, providing details about a preview version for the Xcode IDE from archrival Apple.