Distributed deep neural networks over the cloud, the edge, and end devices

Adrian Colyer, The Morning Paper, Sept 22, 2017
Commentary by Stephen Downes
files/images/ddnn-fig-2ef.jpeg

The next step: "DDNNs partition networks between mobile/embedded devices, cloud (and edge)... What’s new and very interesting here though is the ability to aggregate inputs from multiple devices (e.g., with local sensors) in a single model, and the ability to short-circuit classification at lower levels in the model." Eacj of these two things is equally important. The network is distributed, and the objects described by the network are not the same as the objects escribed by individual members of the network. This article goes into a lot of detail about how they're built and how they function. "By combining multiple viewpoints we can increase the classification accuracy at both the local and cloud level by a substantial margin when compared to the individual accuracy of any device." Original paper (12 page PDF).

Views: 2 today, 374 total (since January 1, 2017).[Direct Link]