Event-driven Tracking with a Particle Filter

The algorithm performs tracking of a circle-shaped target given the event-stream from an event-camera. Two hardware implementations are provided:

@inproceedings{Glover2017a,
address = {Vancouver, Canada},
author = {Glover, Arren and Bartolozzi, Chiara},
booktitle = {IEEE International Conference on Intelligent Robots and Systems},
pages = {3769--3776},
publisher = {IEEE},
title = {{Robust Visual Tracking with a Freely-moving Event Camera}},
year = {2017}
}

CPU

A C++ YARP module that runs the algorithm on a standard CPU.

How to build

1. install YARP
2. install event-driven
3. make sure both YARP and event-driven can be found on the environment path / project path
4. clone this repository (e.g. into ~/projects/particle-filter-tracking)
5. cd ~/projects/particle-filter-tracking && mkdir build && cd build
6. cmake ..
7. make
8. ensure the bin is found in the environment path

How to run

1. yarpserver --write
2. yarpmanager
3. open the provided yarpmanager application (particle-filter-tracking/cpu/app_cpu_tracking.xml)
4. run all
5. open a dataset (example TBA) with the yarpdataplayer
6. connect all
7. play the dataset on the yarpdataplayer

Note: you can substitute the yarpdataplayer with a live feed of an event-camera if it is yarp-compatible. Please contact the authors for more information if you want to make your camera yarp compatible.

SpiNNaker

Python scripts using the "Graph Front End" of the SpiNNaker tool-chain and c-code for running the algorithm on a SpiNNaker board.

• how to compile and run

Setting-up SpiNNaker build (ubuntu)

These instructions assume this repository exists at ~/projects/particle-filter-tracking

1. make directory for SpiNNaker project code
   mkdir /home/user/spinnaker && cd /home/user/spinnaker

2. install cross-compiler from https://developer.arm.com/open-source/gnu-toolchain/gnu-rm/downloads
   Uncompress in /home/user/spinnaker

3. Pull the following repositories from https://github.com/SpiNNakerManchester
   DataSpecification
PACMAN
spalloc
SpiNNakerGraphFrontEnd
spinnaker_tools
spinn_common
SpiNNFrontEndCommon
SpiNNMachine
SpiNNMan
SpiNNStorageHandlers
SpiNNUtils

4. Set-up your ~/.bashrc with the following additional lines
   export SPINN_DIRS=~/spinnaker/spinnaker_tools
export PATH=~/spinnaker/gcc-arm-none-eabi-6-2017-q2-update/bin:$SPINN_DIRS/tools:$PATH
export PERL5LIB=$SPINN_DIRS/tools
export PYTHONPATH=~/projects/particle-filter-tracking/:$PYTHONPATH

   source ~/.bashrc or restart the terminal

5. Apply the bug.patch to repository SpiNNFrontEndCommon
git apply bug.path

6. Run the automatic_make.sh script from /home/user/spinnaker directory

7. cd ~/projects/particle-filter-tracking/spinnaker
make
   At this point the c-code executables should compile. To use the spinnaker toolchain to build the graph and run the algorithm:

Optional: if you want to use an IDE

1. Set-up Python project using your favourite IDE or install PyCharm and make a new project with the contents of ~/projects/particle-filter-tracking/spinnaker/
2. Add all the repositories from step 3. as external libraries of the project.

And/otherwise:

1. Follow the steps under "Configuration" here http://spinnakermanchester.github.io/graph_front_end/4.0.0/SpiNNakerGraphFrontEndInstall.html to set-up your SpiNNaker board.
2. python ~/projects/particle-filter-tracking/spinnaker/pfMain.py