This is a picture of a kitchen demonstrating the capabilities of software by restb.ai. It shows the software identifying features in the kitchen. It identifies stainless steel appliances, a kitchen island, tile floors and natural light.

restb.ai’s software identifies features in a home and generates natural languages for use in all sorts of applications.

We’ve been putting a lot of time into the Hackathon (click here to register) for the iOi Summit, finding good data sources, benefits to the participants, and frameworks they can use. The first sponsor for the Hackathon I wanted to highlight is called restb.ai, which is an example of a company using artificial intelligence and machine learning to create natural language that describes what’s in a picture. Their technology will analyze images you provide and create data sets in the following areas:

  1. Room Type Classifcation
    • Identifies over 30 different room scenes like ‘living room’, ‘kitchen’, ‘bathroom’, etc.
  2. Home Feature Tagging
    • Identifies more than 30 features in the home based on room type. Things like ‘vaulted ceilings’, ‘hardwood floors’, and ‘natural light’.
  3. Exterior Style Classification
    • Has learned 16 different architectural styles for classification purposes.
  4. Logo & Watermark Detection
    • Can identify if an image already has a watermark or logo placement on it.

What’s cool about this type of software is that it creates text data for you on the fly and allows you to do things like populate the machine readable fields for images. It could be used to create listing details, or to create keywords for searches on listings and provide better results.

 

restb.ai will provide access to their API for all hackathon participants to use. We want to thank them for participating! Have you registered for the Hackathon? To find out more and register, click here.