Computer vision and image tagging are making the home shopping experience more accessible and personalized, according to Dominik Pogorzelski, vice president of product for restb.ai, an image tagging software company specifically targeting real estate listing photos.
Pogorzelski defined computer vision as a branch of artificial intelligence (AI) that enables computers to understand the content and context of images or videos. It reads images by classifying them into a category, uses objection detection to pick up what’s in the image and then finds visual similarities to other images.
Using computer vision, a vast set of images can be automatically classified into room types, for example, making the sorting of listing images more streamlined and easier for consumers to view. Consumers could also then look at images of the same room from multiple homes side-by-side, not trying to pair up the specifics manually.
Consumers could also use computer vision technology to see visually similar images. If they see a dream kitchen for example, but the home is way out of their price range, they can see a list of similar kitchens that might be in a more affordable home.
“The real estate industry is sitting on millions and millions of images and doing nothing with it,” Pogorzelski said. “Go out and put your images to work.”
Another major benefit of computer vision is helping make websites more compliant with the Americans with Disabilities Act.
The fast-growing real estate brokerage Compass was recently sued over its listing photos lacking “alt text,” descriptors of the images for blind individuals.
Computer vision would automatically describe a photograph and fill in the alt text field, eliminating these types of lawsuits.
Currently employing this technology would require either a consumer or broker to make a request to computer vision or image tagging application programming interface.