Google AutoML Lets You Train AI Without Needing To Write Complicated Code
Have you wanted to dabble with artificial intelligence but felt overwhelmed by the complexity of it all? Well, good news—Google is rolling out what it calls Cloud AutoML, a suite of machine learning products designed to help developers with limited machine learning experience and expertise to train high quality image recognition models, the company announced today. Basically, it's a way for developers to get their feet wet, rather than jumping into the deep end right off the bat.
"With Cloud AutoML, you can bring your training data to create your own custom vision model with minimum Machine Learning skills required. Start with as little as a few dozen photographic samples and Cloud AutoML will do the rest," Google explains on its AutoML product page.
Cloud AutoML leverages Google's proprietary image recognition technology, and is therefore able to do the heavy lifting that comes with programming machine learning code. To make it easy on developers, Cloud AutoML sports a simple graphical user interface (GUI) to train, evaluate, improve, and deploy machine learning models, all based on the data they provide. It can also be a cost-effective way for companies to leverage machine learning, depending on the task. In that regard, Google has attracted some big names already.
"Cloud AutoML’s technology is helping us build vision models to annotate our products with Disney characters, product categories, and colors. These annotations are being integrated into our search engine to enhance the impact on Guest experience through more relevant search results, expedited discovery, and product recommendations on shopDisney," said Mike White, chief technology officer and senior vice president of Disney Consumer Products and Interactive Media.
Google is also providing additional help. For example, customers who have images but no labels can employ Google's in-house human labelers to review custom instructions and classify images accordingly. Google did not get into the specifics of pricing for that service, but it's a neat option nonetheless.
Cloud AutoML is currently in alpha. Anyone interested in participating can go here and fill out a form.
"With Cloud AutoML, you can bring your training data to create your own custom vision model with minimum Machine Learning skills required. Start with as little as a few dozen photographic samples and Cloud AutoML will do the rest," Google explains on its AutoML product page.
Cloud AutoML leverages Google's proprietary image recognition technology, and is therefore able to do the heavy lifting that comes with programming machine learning code. To make it easy on developers, Cloud AutoML sports a simple graphical user interface (GUI) to train, evaluate, improve, and deploy machine learning models, all based on the data they provide. It can also be a cost-effective way for companies to leverage machine learning, depending on the task. In that regard, Google has attracted some big names already.
"Cloud AutoML’s technology is helping us build vision models to annotate our products with Disney characters, product categories, and colors. These annotations are being integrated into our search engine to enhance the impact on Guest experience through more relevant search results, expedited discovery, and product recommendations on shopDisney," said Mike White, chief technology officer and senior vice president of Disney Consumer Products and Interactive Media.
Google is also providing additional help. For example, customers who have images but no labels can employ Google's in-house human labelers to review custom instructions and classify images accordingly. Google did not get into the specifics of pricing for that service, but it's a neat option nonetheless.
Cloud AutoML is currently in alpha. Anyone interested in participating can go here and fill out a form.
