Jernej Kavka

Getting Started with Machine Learning using ML.NET

Want to get started with machine learning but don't know where to start? Have you got an Excel spreadsheet, SQL Database or CSV lying around and wondering if you can use it to experiment with Machine Learning?

In this workshop, we'll start from a CSV exported by a service, and go all the way to an application that uses Machine Learning to make clever decisions.

We will cover:
1. What does a developer need to know about Machine Learning?
2. How does ML.NET help getting started with ML?
3. Quickly prototype a solution with ML.NET Model Builder
4. Improve solution with simple data science rules
5. Integrate a machine learning solution into your application
6. Continuously improving machine learning model and updating applications

From Paper to Power using Azure Form Recognizer

Are you still collecting your feedback on paper handouts? Red, yellow, green feedback is faster, but what do you do with that? It’s so time-consuming to go through the feedback forms afterward and maybe you just don’t bother. What if you could collect feedback on a custom form, collate the data and get actionable results before your coffee gets cold?

Join me on the journey from waiting weeks to get feedback from the user group talks to have the results in less than an hour for each event. You’ll see how you can use Form Recognizer to parse the data straight from the page, what you can and can’t do right now, as well as how you can leverage other Cognitive Services to get more details from the user feedback forms!

Making unit tests simple again with .Net Core and EF Core

Unit testing can be hard especially when databases are involved.
In this talk, JK will teach you how to include EF Core to simplify testing of “Will this actually work on a real DB” while keeping the tests self-contained and repeatable.

Also walk away knowing how cool the combination of EF Core in-memory and the SQLite DB provider is.

Real-time Face Recognition With Microsoft Cognitive Services

This session is all about Microsoft Face API in practice!
JK the man behind sswSophie.com will cover what it means to use Microsoft Cognitive Services in real-time, why offline detection libraries are essential and why you should use infrared and depth cameras like Kinect and Intel RealSense in your applications.

SSW TV - Real-time Face Recognition With Microsoft Cognitive Services

JK, the man behind www.sswSophie.com, will cover what it means to use Microsoft Cognitive Services in real-time, why offline detection libraries are essential and why you should use infrared and depth cameras like Kinect and Intel RealSense in your applications.

YouTube video: https://www.youtube.com/watch?v=KCSyRO0KotA

Delivered in January 2019 in Melbourne, Canberra, and Sydney.

Here is the demo project built with C#, OpenCV and Cognitive Services:

https://github.com/jernejk/RealTimeFa...

Here are some of the frameworks for offline face detection:

C# ↴
- UWP
https://msdn.microsoft.com/en-us/maga...
- OpenCV
https://github.com/shimat/opencvsharp

JavaScript ↴
- TensorFlow.js
https://js.tensorflow.org/
- BRFv4 https://github.com/Tastenkunst/brfv4_...

--- About the presenter ---

With around 10 years of experience in software engineering, Jernej has worked on full-stack .NET development, mobile applications, and Microsoft Cognitive Services. He worked for some of Australia's largest corporations, with great customer satisfaction.

Machine Learning simplified for Developers with ML.NET

Do you want to try machine learning, but don't want to invest too much time learning a new programming language or some other complicated API?

Microsoft recently launched ML.NET 1.4 which is a great entry point for .NET developers and to gain experience building something with Machine Learning.

With the recent release of ML.NET Model Builder, we can create machine learning models by attempting to import raw data first and over time curate the data, to get better results.

JK will show you how ML.NET works, how to leverage Model Builder, experiment with training data and what to watch out for when building models.