Microsoft Corp. announced the general availability of several new services to Azure, including Azure Custom Vision, a feature of Azure’s Cognitive Services.
This service, which is powered by machine learning, allows users to build, improve, and implement image classifying capabilities; which tags images with labels by detecting visual characteristics.
The service will be customizable so that the user can train exactly what visual indicators are important to the work they are using it for.
Users of the Azure Custom Vision service train algorithmic models by uploading labeled images.
“Customers can train their own custom image classifiers and object detectors,” said Tina Coll, the product marketing manager at Microsoft Corp. “For example, a company could choose to detect their own logo in the video of a sports event to track the impact of their advertising or a student might want to count the number of animals passing in front of a nature camera.”
Another intriguing aspect to this service is that it can be uploaded to devices that use iOS (in CoreML), Android (in TensorFlow) and other devices that use edge computing, and can be run offline in real-time.
This could have serious ramifications for remote field workers; allowing them to use these services to conduct their tasks in areas that may have limited or no connection.
Coll provided examples of those it could benefit, including workers on oil rigs who may need cognitive services for their job tasks.
One company who already uses this service, in conjunction with Azure video analytics, is Minsur in Peru; the largest tin mine in the western hemisphere. Cognitive Vision ensures that the water it uses in the mineral extraction process is still reusable and not tainted by the treatment process.
Custom Vision is not yet available from Azure’s Canadian data centres, but has been released across most of the U.S.