Baidu launches 'no code' AI model training platform

September 05, 2018 // By Rich Pell
Chinese internet giant Baidu (Beijing, China) has launched an AI platform designed to make building custom machine learning models easier that does not require algorithmic programming.

Geared toward small and medium businesses, the platform - called EZDL - is said to allow everyone "to access AI in the most convenient and equitable way." The online tool allows users to create and deploy AI models without requiring any coding skills or needing large amounts of data.

"EZDL is a service platform that allows users to build custom machine learning models with a drag-and-drop interface," says Yongkang Xie, Tech Lead of Baidu EZDL. "It takes only four steps to train a deep learning model, built specifically for your unique business needs."

“Even if you have had no exposure to programming, you can quickly build models on this platform with zero barriers," says Xie. "EZDL can help companies with limited AI experts and computing resources to quickly and efficiently complete deep learning training and deployment with only a small amount of data."

EZDL models, says the company, can be built with smaller training sets – often just tens or hundreds of examples - compared to the thousands or even millions typical of other models. This also results in significantly reduced training times.

EZDL is offered as featuring four key characteristics: it is easy to use, fast, high performance, and highly secure. Extremely friendly to entry-level AI developers, it opens up capabilities in image classification, object detection, and sound classification.

The company presents several example case studies for each of the above models. For the image classification model, it cites examples using it for home decoration, industrial quality inspection to identify defective products, pavement damage monitoring, and image recognition of Chinese herbal medicine, wild birds, and frogs.

The object detection model, says the company, has been used for retail inspection and medical detection for cell counting. Finally, the sound classification model has been used for security monitoring and real-time detection of dangerous situation, as well as voice recognition of different species and individuals.

The platform is part of the company's


s

Vous êtes certain ?

Si vous désactivez les cookies, vous ne pouvez plus naviguer sur le site.

Vous allez être rediriger vers Google.