AWS Releases DeepLens Camera for Devs with Built-In Machine Learning
Amazon Web Services (AWS) has begun shipping DeepLens, a camera with built-in machine learning capabilities that it first announced at last November's re:Invent conference.
The devices cost $249 and are available to just U.S.-based customers initially, according to a blog post earlier this month by AWS evangelist Jeff Barr, though shipping to "additional destinations" is coming.
DeepLens lets developers take photos of real-world objects and easily incorporate them into their machine learning projects, even if they don't have extensive experience in machine learning. "AWS DeepLens allows developers of all skill levels to get started with deep learning in less than 10 minutes by providing sample projects with practical, hands-on examples which can start running with a single click," AWS claims on the DeepLens Web site.
DeepLens comes with deep learning models built-in, or developers can import their own models from the Amazon SageMaker machine learning platform. In the first public demo of DeepLens at re:Invent, AWS AI executive Matt Wood showed how SageMaker can be used to build a machine learning model that gives music recommendations based on users' past song selections. Wood then fed that SageMaker model into DeepLens and used the camera to create "reviews" of particular albums by simply taking pictures of his face next to the album covers. DeepLens was able identify Wood's sentiment for each album based on his expression, and assigned reviews accordingly.
DeepLens also integrates with other AWS services, such as AI solutions like Polly and Rekognition and the Lambda programming environment.
The DeepLens device itself is compact, measuring roughly 140 millimeters tall, 47 millimeters wide and 94 millimeters deep. The 4MP camera supports 1080p video. It also has Wi-Fi connectivity, supports a 2-D microphone array and comes with two USB 2.0 ports.
Inside, DeepLens is powered by an Intel Atom processor, runs Ubuntu OS-16.04 LTS and has capacity for 8GB of RAM with 16GB of expandable memory.
According to Barr, AWS has added several new features to DeepLens since its preview phase, including support for the Caffe and TensorFlow deep learning frameworks, support for more MXNet layers, integration with Amazon Kinesis Video Streams and a new sample project that detects head positions.