RPA Computer Vision Masterclass with Neurolabs

This live video with Radu from Neurolabs teaches you the basics of Computer Vision and how to implement it into your RPA workflows. The Computer Vision from Neurolabs lets you easily detect objects from images with the use of drag and drop no-code software. You can connect your RPA workflow to the Computer Vision with an API connection (http request).

Links:

Books [affiliate]:

You could also watch:

0:00 Intro
This video is about Neurolabs and Computer Vision. Computer Vision is a part of Intelligent Automation, which you need to learn as an RPA Developer. Intelligent Automation is how we automate white collar office jobs. Each of the pillars represents one thing that the thinking office worker is performing. Do (RPA), Learn (Machine Learning, Big Data), Talk and Listen (Chatbots and Sentiment Analysis) and Seeing (Computervision, OCR). Radu Buciu will present the case. Neurolabs is very easy to use, since it’s drag and drop.

2:06 Radu Buciu, Neurolabs
Radu Buciu is Project Manager in Neurolabs. Neurolabs is a Computer Vision start-up with a clear vision to democratize Computer Vision. Radu himself is an experienced RPA developer. Can you master Computer Vision/Machine Learning with your RPA workflows, you can get ahead of the competition. Neurolabs won the Start-Up Category at UiPath Automation Awards in Turkey 2020.

6:00 Quick overview of Neurolabs
Neurolabs is a way for developers with no machine learning experience to train models. The data is generated synthetically, which makes the process a lot faster and cheaper. The dataset is generated (5000 images) and then we can train the model.

8:15 Computer Vision - The steps to follow
Get Data, Clean and Label the Data, Train Model, Test Data and Improve.

9:37 Intelligent Automation

10:13 Object Detection in Computer Vision
Take a lot of pictures, manually label them and train deep learning models. This is inefficient and the images are limited.

11:42 Why use synthetically generated data
At Neurolabs we use synthetically generated data to be more efficient. It will be very easy to generate the pictures instead of having people going around and taking them.

13:40 3D Models
We can easily add our own models at the Neurolabs platform (drag and drop). We can find 3D models at turbosquid or take them ourselves with an iPhone.

22:50 UiPath RPA workflow with Neurolabs Computer Vision
Radu creates a UiPath RPA workflow that reads images from a live webcam of a parking lot, then processes them via a REST API connection. The results are both a picture with boundary boxes and a JSON with the classes and confidence levels.

30:04 Get Neurolabs
You can get a free trial of Neurolabs, if you have a great use case to apply.

Connect with me:

#rpa #computervision #automation

[UiPathHacks]

very exciting, guys. Congrats! Waiting to get some hands-on experience with it :)

[arazdavlund]

Thanks Anders, it's very interesting.

[SteenBendtsen]

Really interesting, I look out for Neurolabs

[djurokukic]

Is this tool meant to be used for reading text, numbers, punctuation marks etc?