Qualcomm is to release the Qualcomm Snapdragon Neural Processing Engine, a deep learning software development kit (SDK) for devices powered by Qualcomm Snapdragon 820 processors.
The SDK is powered by Qualcomm’s Zeroth Machine Intelligence Platform and has been optimised to use Snapdragon’s heterogeneous computing capabilities. The SDK offers OEMs a powerful, energy efficient platform that can offer intuitive and engaging deep learning-driven experience on device.
Qualcomm claims that it is the first mobile SOC manufacturer to offer a deep learning toolkit optimised for mobile devices.
With this SDK, OEMs can work on their own neural network models on their devices including smartphones, tablets, security cameras, automobiles and drones without the need to connect to any cloud.
Some of the most common deep learning user experiences that can be realised include face recognition, gesture recognition, text recognition, object tracking and avoidance, scene detection and also natural language processing.
Zeroth Technology presently drives visual intelligence software like Snapdragon Scene Detect and malware detection software found in Snapdragon Smart Protect.
Qualcomm Technologies product management director Gary Brotman said: "The demand for untethered, mobile learning-driven user experiences is increasing rapidly and Qualcomm Technologies’ customers have been clamoring for tools to help them realise their product visions.
"With the introduction of the new Snapdragon Neural Processing Engine SDK, we are making it possible for myriad sectors, including mobile, IoT and automotive to harnesses the power of Qualcomm Snapdragon 820 and make high-performance, power efficient on-device deep learning a reality."
Snapdragon’s Neural Processing Engine SDK features an accelerated runtime for on-device execution of convulutional and recurrent neural networks on Snapdragon 820 cores. It can support common deep learning models.
The Neural Processing Engine SDK has a lightweight and flexible platform designed to utilise Snapdragon’s heterogeneous cores to offer optimal performance and reduce power consumption.
Qualcomm also says that the neural network processing technology can also be utilised in several other industries such as healthcare, automotive, security, imaging and other industries. Companies can user their own network and hand-held devices.
NAUTO chief technology officer Frederick Soo said: "The Qualcomm Snapdragon Neural Processing Engine SDK helps us bring deep learning to our connected camera and smart cloud network, and that helps us present a full, contextual picture of the driving environment to our commercial automotive fleet and auto insurance customers.
"The Neural Processing Engine SDK means we can quickly deploy our proprietary deep learning algorithms to our Snapdragon-based connected camera devices in the field, which can detect driver distraction and help prevent auto accidents.
"That same deep learning helps NAUTO’s system deliver valuable information that helps our customers reduce liability and drive more safely and efficiently."