This project comprises of two main parts. A server that has the NCS connected to it, and a chrome extension that allows a user to interact with the NCS devices. The user should be able to acquire access to a device, load a network, and run the inferencing program either on images or on video.
For example, if you had a neural net which can generate captions for video, or track moving object, you could use this project run the net on any video you are watching and have the results displayed as an overlay to the video in real time .
Some work on this was started during my internship at Intel last summer, and I would like to take it from a rudimentary proof of concept to a full fledged version that implements all of the above mentioned features and has a layer of security.