Google Teaches AI To Build AI For Beginners, Starts With Images

Teams at Google have begun to use artificial intelligence (AI) to develop a suite of machine-learning products that enable developers with limited skills to build their own custom system based on AI.

The service, Cloud AutoML, uses machine learning to automatically build and train learning algorithms. The first one -- Vision -- was announced Wednesday, and recognizes objects in images. The model provides a simple graphical user interface that enables a layperson to specify data, then turns that data into a model customized for specific needs. There is also a faster turnaround for production-ready products, of as little as a day.

The platform addresses the lack of talent, knowledge and expertise in AI. "Only a handful of businesses in the world have access to the talent and budgets needed to fully appreciate the advancements of ML and AI," wrote Fei-Fei Li, chief scientist of Cloud AI at Google, in a blog post. "There’s a very limited number of people that can create advanced machine learning models. And if you’re one of the companies that has access to ML/AI engineers, you still have to manage the time-intensive and complicated process of building your own custom ML model." 

Several companies have been testing AutoML for the past few months. A Vision model built into Disney's search engine on its website aims to improve the ability to search and discover products in the Disney store.

Urban Outfitters has begun to use Cloud AutoML to address attribution, according to Alan Rosenwinkel, data scientist at URBN. He explains in the blog post how the company's using the technology to automate the product attribution process by recognizing nuanced characteristics in clothing like patterns and neckline styles.

The products are being developed with help from Google Brain engineers and others across Google.

Next story loading loading..