Pokedex app using Deep Learning
Sorry for the quality of the video, my Raspberry Pi is pretty slow and i wasn't able to record the screen.. so i had to improvise. Hope it's still pretty clear what i'm doing there and how the Pokedex app is working.
This is a Virtual Pokedex app made in Python on a Raspberry Pi with Deep Learning that is able to determine pretty good which Pokemon it is from a list of 5 pokemons (for now): Bulbasaur, Gastly, Hitmonchan, Piplup and Togepi. I used an Intel Movidius Neural Compute Stick because the Raspberry itself is pretty slow and i needed a lil' bit of extra speed :) I used Tensorflow, Inception V3 for the graph file