Left: A sample cascade detection. Right (six images): White pixels indicate the locations where scores were computed, at the scale of the detection (left), for each of the six part filters. Note how scores are not computed at most locations, yielding a fast detection.

Overview

This is an implementation of our star-cascade algorithm for object detection with deformable part models. This algorithm is described in [1]. See [2] for more general information about our object detection system.

By building cascade detectors for our deformable part models we obtain an average detection time speedup of roughly 14x on the PASCAL 2007 dataset with almost no effect on AP scores. By allowing for a reduction in recall (i.e., increasing the global detection threshold), we can obtain speedup factors as high as 40x for some classes with almost no loss in precision. (All timing comparisons are between single-threaded implementations of both algorithms.)

This package contains object detection and cascade threshold learning code. It also contains pre-computed statistics needed to select thresholds for the models included in the release of our object detection system. These statistics were computed on the PASCAL 2006, 2007, 2009, and INRIA Person datasets. In each case, the statistics were computed on the same data that the detectors were trained on. In principle, thresholds selected from this data could be too aggressive due to over-fitting, however we have not observed this problem in practice.

The system is implemented in MATLAB, with the core algorithm written in C++ for efficiency reasons. The software was tested on several versions of Linux and Mac OS X using MATLAB versions R2009b and R2010a. There may be compatibility issues with other versions of MATLAB and with other platforms.

For questions regarding the source code please contact Ross Girshick at r...@cs.uchicago.edu (click the "..." to reveal the email address).

Download

This package is an add-on for our object detection system. You must install that system first in order to use the cascade code.

Get the star-cascade source code. Updated: Feb 28, 2012 (voc-release4.01 compatibility fix)

This project is now hosted on github.

Installation

  1. Download the package and extract it in your voc-release4 directory.
    $ cd /path/to/voc-release4
    $ tar zxvf voc-release4-star-cascade.tgz
  2. Open the README file and follow the detailed installation instructions within.

References

[1] P. Felzenszwalb, R. Girshick, D. McAllester.
Cascade Object Detection with Deformable Part Models.
To appear in the Proceedings of the IEEE CVPR 2010.
pdf

[2] P. Felzenszwalb, R. Girshick, D. McAllester, D. Ramanan.
Object Detection with Discriminatively Trained Part Based Models.
To appear in the IEEE Transactions on Pattern Analysis and Machine Intelligence.
pdf