What is Deep Learning

In Rubiscape, RubiAI deals with a simplified study of Deep Learning, applied to multiple areas like video analytics and imagery. RubiAI employs the Deep Learning techniques of image recognition to enable

• Computer Vision Analytics
• Image Processing
• Enhancement and Transformation, and
• Object Analysis

RubiAI provides Convolutional Neural Network (CNN) models for instructing computers to understand and interpret any visual form of information. Using RubiAI, you can perform interactive image and video data exploration and automatically tune the analysis model parameters. You can also address issues in images, signals, and videos.
RubiAI is the right tool for AI experts like AI-ML Engineers and Computer Vision Engineers since it contains pre-built models for image and video analysis.

Image Analysis in Deep Learning

RubiAI provides two pre-built features for image recognition and analysis. These are

• Face Detection
• Face Verification
• Image Capturing 
• Number Plate Detection

Both the features contain pre-trained models for performing image analysis.
To access these pre-built features, in the task pane, click Deep Learning.

To perform image analysis, first, the images are added to a particular folder in Rubiscape using RubiPython script. You need to write a script to access and read external images or images from the AWS or S3 databases, add them to the folder, and then use them for analysis.
Next, you need to create a Custom Output Variable in RubiPython and store the variables in a dictionary in the form of keys.
For example, in the case of number plate detection, we create output variables for the vehicle images and the corresponding number plates.
For more information, refer to RubiPython.

Image analysis is employed in AI-enabled CCTV cameras used by traffic control authorities to detect vehicles that jump traffic signals or break traffic and driving regulations.