Amazon is presently sending the $250 DeepLens camera it divulged in November to enable designers to make AI picture acknowledgment applications and models. In spite of the fact that it looks pretty disappointing, there’s a considerable measure of intensity in the smallish body. The camera 4-megapixel camera can shoot 1080p video, and the smaller than expected PC beneath it packs 8GB of RAM and an Atom processor with 106 gigaflops of execution – effortlessly enough to run visual AI models disconnected.
You get the typical cluster of I/O ports and remote alternatives vital for application improvement, including Micro HDMI, USB 2.0, a 2D amplifier exhibit and double band WiFi. Amazon said you don’t really require machine learning or PC vision encounter, and demonstrated it by running a hackathon, with most groups never having done any AI programming.
DeepLens underpins some of Amazon’s AI administrations, including Greengrass, TensorFlow and SageMaker, its most recent machine learning device (beneath). It accompanies venture layouts that can recognize 20 protests, a face location module, and another that can perceive around 30 activities like guitar playing (Amazon has an advantageous guide here that makes everything look quite straightforward.) As said, you would now be able to now arrange the camera today in the US (it should come somewhere else later) for $250.