Is there any working with deploying NN on embedded devices

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

I am working on Raspberry Pi and multiple cameras for trajectory tracking. Is there anyone already working in this area?

1 Like
2

Probably yes, given the popularity of the RPi series of computers.

The general term is “Edge AI”.

2 Likes
3

In addition to Tom’s more wide ranging reference, Prof Ng does discuss some of the issues around deploying models on memory and cpu constrained devices here in DLS. In particular, see the section in DLS Course 4 Week 2 starting with the lecture on Network in Network and 1 x 1 Convolutions through the end of Week 2 where he discusses MobilNet and EfficientNet. Those are architectures that were specifically developed to be deployable in constrained environments. Of course Moore’s Law doesn’t sit still, so what constitutes a “constrained environment” continues to change over time. :nerd_face:

1 Like