Large Scale Metric Learning from Equivalence Constraints*
Martin Köstinger, Martin Hirzer, Paul Wohlhart, Peter M. Roth, Horst Bischof
Institute for Computer Graphics and Vision, Graz University of Technology
* The work was supported by the Austrian Science Foundation (FWF) project Advanced Learning for Tracking and Detection in Medical Workflow Analysis (I535-N23) and by the Austrian Research Promotion Agency (FFG) project SHARE in the IV2Splus program.
This page briefly describes our work on large-scale similarity computation.
- the related CVPR 2012 paper
- the matlab code
- competing algorithms
- all evaluated datasets with extracted features
Prerequisites to run the code:
- Linux (x64) and Windows (x32,x64) are tested, should also work on other platforms.
Datasets and Results
- Face verification on LFW (samples,link)  and PubFig (samples,link) 
- Person re-identification on the VIPeR (samples,link)  dataset
- Matching unseend objects on the LEAR ToyCars (samples,link)  dataset
To check out the result figures where we benchmark our method (KISSME) to related work click here (LFW, ToyCars, VIPeR, PubFig). The biggest benefit is the speedup in training compared to related approaches.
The best thing is to check out the inline documentation of the matlab files. For all experiments a main file (LFW, ToyCars, VIPeR, PubFig) describes the used features and database. Our method is described in LearnAlgoKISSME.m, in particular in the learnPairWise method.
Instructions for reproducing experiments
In order to run the experiments presented in this paper, we assume basic knowlege of Matlab. The code is tested under Linux (x64) and Windows (x32,x64), should also work on other platforms with minor modifications. Reproducing the results of our method should not take more than five minutes, including installation.
To download and unpack the needed files run the following code snipet in Matlab. Alternatively, click the links; download and unpack the archives to a directory of your choice.
The provided features cover everything that is needed to reproduce the experiments. If you want to use the original extracted features (before PCA compression) download the kissme_features_full.zip (887.11 MB) archive.
The experiments described in the paper benchmark our method (KISSME) to other metric learning methods (LMNN,ITML,LDML,SVMs). Due to different licenses these are not pre-installed by default. If you agree to these install the code with the following matlab snipet. For details click on the link below.
Quick Start Running Experiments
Change the directory to KISSME, workflows, CVPR.
Pick the experiment of your choice and run the according script, i.e. demo_viper.m.
To run all experiments ...
Note: For LFW and PubFig only KISSME is enabled per default. For some of the other algorithms it takes quite long to complete. If you want to train all installed learning algorithms uncomment the respective code. Check the inline comments for details, e.g. in demo_lfw_sift.m.
LicenseThe toolbox code is licensed under the BSD 3-Clause License ("modified BSD license"). If you use the code, i.e. our algorithm in a scientific publication please cite this paper. For the provided data please check the included copyright notice as it is partly based on other data.
 L. M. Bregman. The Relaxation Method of Finding the Common Point of Convex Sets and Its Application to the Solution of Problems in Convex Programming. USSR Computational Mathematics and Mathematical Physics, 7:200-217, 1967.
 M. Farenzena, L. Bazzani, A. Perina, V. Murino, and M. Cristani. Person re-identification by symmetry-driven accumulation of local features. In Proc. IEEE Intern. Conf. on Computer Vision and Pattern Recognition, 2010.
 D. Gray, S. Brennan, and H. Tao. Evaluating appearance models for recongnition, reacquisition and tracking. In Proc. IEEE Intern. Workshop on Performance Evaluation of Tracking and Surveillance, 2007.
 S. C. Hoi, W. Liu, M. R. Lyu, and W.-Y. Ma. Learning distance metrics with contextual constraints for image retrieval. In Proc. IEEE Intern. Conf. on Computer Vision and Pattern Recognition, 2006.
 G. B. Huang, M. Ramesh, T. Berg, and E. Learned-Miller. Labeled faces in the wild: A database for studying face recognition in unconstrained environments. Technical Report 07-49, University of Massachusetts, Amherst, 2007.