亚马逊发布深度学习相机

Developing AI* systems just got a lot easier, thanks to Amazon’s DeepLens camera.

The US$250 gadget can recognize objects and faces and automatically label what actions people are taking, which programmers can use to develop their own machine learning apps and hardware.

Putting the affordable technology into the hands of smaller companies could fuel a boom in smart systems, like the retail giant’s own Alexa assistant.

DeepLens was developed by Amazon Web Services (AWS), a subsidiary of the Seattle firm that provides an on-demand cloud computing platform to businesses and consumers.

It comes packed with a suite of AI tools that will let developers with little experience of machine learning design and test out their own creations. The device does so without needing to connect to AWS’s servers, speeding up the process of testing.

DeepLens is already available via Amazon’s US site, although it is not known when it will be released globally.

DeepLens uses a four megapixel* camera to capture 1080P high definition video, as well as a microphone to record audio.

It is powered by an Intel Atom processor and 8GB of memory, which can run the frames of incoming video through on-board deep learning models.

These models can recognize cats and dogs, faces, a wide array of household and everyday objects, and more than 30 motions and actions. This includes brushing teeth, applying lipstick and playing the guitar. The camera can connect to networks and devices through Wi-Fi, USB and micro HDMI* ports. (SD-Agencies)