Personal tools
You are here: Home CG seminar 2018 Real-time Machine Learning using Core-Sets: Autonomous Toy-Drones for Rami Levy
« October 2018 »
October
SuMoTuWeThFrSa
123456
78910111213
14151617181920
21222324252627
28293031
Log in


Forgot your password?
 

Real-time Machine Learning using Core-Sets: Autonomous Toy-Drones for Rami Levy

Wednesday, November 22nd, 2017, 16:10

Schreiber 309

underline

Real-time Machine Learning using Core-Sets: Autonomous Toy-Drones for Rami Levy

Dan Feldman, Haifa University

Abstract:

A coreset (or core-set) of a dataset is its semantic
compression with respect to a set of classifiers, such that learning
the (small) coreset provably yields an approximate classifier of the original (full) dataset. 
However, we are not aware of real-time systems that compute coresets in
a rate of thousands of points per second.
 
We suggest a framework to run geometric optimization on 
real-time systems with provable guarantees using coresets. This is by maintaining coresets for kinematic (moving) set of points, and run algorithms on the small coresets, instead of the original input points in real-time and using weak devices. 
 
This enabled us to implement a low-cost (< $50) autonomous toy-drones with only a single camera on-board that guides guests to a desired room (in a hospital, mall, hotel, museum, etc.), and the first assistant swarm of drones inside malls (Grand Canyon in Haifa) and supermarkets (Rami Levy Hashikma Marketing).
 
A joint work with Soliman Nasser and Ibrahim Jubran and results from ICML'17




Document Actions