DeepLens offers product development with artificial intelligence and machine learning

Amazon Web Services (AWS) has presented its wireless camcorder, completely programmable and developer-oriented, which has had the collaboration of Intel.

The DeepLens wireless camera, presented by Andy Jassy, CEO of AWS during his presentation at the annual conference re:Company Invent, recently held in Las Vegas (EU), arises from the collaboration with Intel to “offer creators of any level of experience an optimal and necessary tool to design and create artificial intelligence products (IA) y machine learning”.

According to this company, both artificial intelligence and machine learning They will be the drivers of a new generation of smart industries, including trade, factories and many other applications “that will make our lives easier through more fluid and natural interactions with these devices”.

AWS and Intel's technology collaboration reinforces Intel's commitment to providing developers with advanced tools to create AI products and machine learning, just as it did recently with the presentation of Speech enabling developer, the developer voice training kit, providing a complete audio solution that facilitates voice control in the open field and allows third-party developers to accelerate the design of consumer products that integrate the Alexa voice service.

DeepLens combines high throughput with an easy-to-use user interface to support model deployment in the cloud.

This device is based on an Intel Atom X5 processor, with integrated graphics to support object detection and recognition, in addition to using software tools and libraries deep learning (including Intel Math Kernel Library), improved by this company to run computer vision models in real time directly on the device.

As AWS and Intel point out, Developers can start designing and building AI products and machine learning in minutes using pre-configured DeepLens frames. At the moment, The Apache MXNet environment is supported and it is expected that in early 2018 capabilities for Tensorflow and Caffe2 will be added.