Google Introduces AutoML Video and AutoML Tables for Structured Data
People with no coding experience can use AutoML Tables to develop custom AI models by utilising structured tabular datasets
Google announced the release of AutoML Video and AutoML Tables for structured data as updates for AutoML after its launch nearly a year ago. These two new classes for suite of services by Google can automate the development of automated artificial intelligence (AI) systems.
Google has introduced the Cloud AutoML for the development of custom AI models in January last year.
People with no coding experience can use AutoML Tables to develop custom AI models by utilising structured tabular datasets. These tables can collect data from the BigQuery data warehouse of Google Cloud Platform (GCP) and other storage providers.
Able to use natural language
Similar to AutoML Video Intelligence service, AutoML Video was first introduced in late 2017. The AutoML Video will be able to utilise natural language and translation to transcribe conversations, and computer vision to identify things such as explicit content and scene changes. This can also be used for developing custom classification models that can serve the unique requirements of customers.
Speaking on the development, Rajen Sheth, Senior Director – Product Management, Google Cloud, said, “We’re also seeing in most industries things like demand forecasting, all the way through to things like price optimisation. All of those are structured data problems and things AutoML Tables can be applied to.”
“Objection detection for AutoML Video is coming soon, he added.
In July last year, Google released drag-and-drop tool of AutoML Vision for training visual systems in public beta. In addition, it introduced AutoML Translate and AutoML Natural Language.
source : http://electronicsofthings.com