New Zealand vertical Homes.co.nz has signed a deal with Restb.ai to implement the company’s image recognition technology on its website. Photos uploaded to Homes.co.nz will now automatically be classified according to room type and Kiwis will be able to search properties by rooms.
The site has launched the new feature as Homes Inspiration under a business agreement with New Zealand’s largest paint supplier Resene. “Homes Inspiration is a new platform harnessing the power of computer vision technology to automatically classify the room type of real estate listing photos,” Homes.co.nz chief data scientist Tom Lintern explained.
According to a Restb.ai press statement, Homes.co.nz now offers a growing collection of more than 200,000 images in categories like Kitchens, Bathrooms, Living rooms, Bedrooms, and Laundries. “Property photos have the ability to evoke emotion in a unique way. With Homes Inspiration, Homes.co.nz is leveraging that emotion to help match buyers with the photos they want to see,” Restb.ai CEO Angel Esteban said.
Homes.co.nz is a hybrid between a classified and data site. “They purchased all the data about properties from local councils and mapped them on their site,” AIM Group analyst Angela Hawksford explains. “Whenever an agent lists a property on the open market —- either with Homes.co.nz or another NZ property site — they mark it as ‘on the market’ and show the active listing. They also show previously sold properties.”
While Restb.ai and Homes.co.nz claim this technology is a New Zealand first, there’s nothing new about using artificial intelligence and machine learning in classifieds. Restb.ai already counts German ImmobilienScout24 and Spanish Idealista among its real estate classified clients.
Prosus-owned OLX introduced its image recognition technology in 2016. EBay-owned horizontal Marktplaats, in the Netherlands, has been using image recognition for its listings since October 2017. Carsales in Australia released what it called a ‘world first’ in image recognition software for auto sites in June 2017. Called Cyclops, the software uses AI to select and assign angles to images uploaded to the site by users, to help showcase various aspects of the vehicle to buyers.
In October 2017, Lagou — a leading recruitment platform for the Chinese tech sector — started using an AI- and facial recognition-based identity authentication system to push back against fake job postings. The software triangulates between a live image of the user’s face, the uploaded photo, and the official photograph found on the state-mandated identity card.
Last year, horizontal LetGo introduced its Reveal feature. The machine learning tool automatically suggests a price, title, and category for images loaded to the site.
The AIM Group took a deep-dive look at the companies using AI, how they’re using it, and why, in July (AIMR, Vol. 20 No. 16 – July 9, 2019 ).
Barcelona-based Restb.ai offers AI technology that identifies, classifies and tags features and settings in online property images. It also offers photo compliance and watermark detection tools. Its benefits include boosting Google rankings by correctly tagging photos with search engine optimization keywords and enabling voice search, thanks to its tags and descriptions. The start-up won €1.7 million ($1.97 million U.S.) in funding from the European Innovation Council — beating some 1,600 other European start-ups — in August last year.