Awesome
DeepPed: Deep Convolutional Neural Networks for Pedestrian Detection
Created by Denis Tomè, Federico Monti, Luca Baroffio and Luca Bondi.
Introduction
DeepPed is a state-of-the-art pedestrian detector that extends R-CNN work done by Girshick et al. combining region proposals with rich features computed by a convolutional neural network. This method achieves 19.90% log-average-miss-rate on the Caltech Pedestrian Dataset.
DeepPed is described in an arXiv tech report and will appear in Elsevier Journal of Signal Processing.
Citing R-CNN
If you find R-CNN useful in your research, please consider citing:
@article{tome2015Deep,
author = {Tomè, Denis and Monti, Federico and Baroffio, Luca and Bondi, Luca and Tagliasacchi, Marco and Tubaro, Stefano},
title = {Deep convolutional neural networks for pedestrian detection},
journal = {arXiv preprint arXiv:1510.03608},
year = {2015}
}
}
License
DeepPed is released under the Simplified BSD License (refer to the LICENSE file for details).
Installing R-CNN
- Prerequisites
- MATLAB (tested with 2015a on 64-bit Linux)
- Caffe's prerequisites
- Install Caffe and R-CNN
- Download Caffe (version described in R-CNN instructions)
- Download R-CNN and follow the instructions
- Install DeepPed
- Change into the R-CNN source code directory:
cd rcnn
- Get the DeepPed source code by cloning the repository:
git clone https://github.com/DenisTome/DeepPed.git
- Get the Piotr's Image & Video Matlab Toolbox by cloning the repository:
git clone https://github.com/pdollar/toolbox.git
- From the
R-CNN
folder, run the model fetch script:./DeepPed/fetch_models.sh
. - Open the
startup.m
matlab file, adding the two commandsaddpath(genpath('DeepPed'));
andaddpath(genpath('toolbox'));
at the end of the file.
Running DeepPed on an image
- Change to where you installed R-CNN:
cd rcnn
. - Start MATLAB
matlab
.
- Important: if you don't see the message
R-CNN startup done
when MATLAB starts, then you probably didn't start MATLAB inrcnn
directory.
- Run the demo:
>> deepPed_demo