IsoGD baseline

This is the source code of the baseline method of IsoGD, a large-scale isolated gesture dataset. Baseline method: MFSK features -> Kmeans -> SVM

ChaLearn Challenge and Data downloading

The ChaLearn LAP Large-scale Isolated Gesture Recognition Challenge is in progress, please feel free to participate!
You can gain access to the ChaLearn LAP IsoGD dataset from http://www.cbsr.ia.ac.cn/users/jwan/database/isogd.html.

Notes

This code was tested on Ubuntu 14.04 OS, with MATLAB 2013b and Python 2.7. There is a compiled MFSK binary program for Ubuntu 14.04.
Please double check the paths in code before your run it.

##Steps in detail of baseline method:

####Step 1. Extract MFSK Features. For both train_list.txt and valid_list.txt use isogd/batch_MFSK_feature_isogd.m program to extract and save MFSK for all videos.

####Step 2. Calculate Kmeans Center and Save Hists Excutes joewan_upload_V2/joewan_code/iso_cal_center_and_get_hists.m program to do all the works. It first randomly selects 20 videos for each class in train set, and uses features of these videos to calculate the center. When the center is ready, it will calculate and save the hist using the MFSK features of each video in both train set and validation set.

####Step 3. Use Lib-SVM to Train SVM Model and Predict the Label of Validation Hists. First scale both train.hist and valid.hist to range [-1, 1], using commandsvm-scale. Then train svm with linear kernel type svm-train -t 0. At last, use trained model to generate labels for valid.hist.

####Step 4. Format svm prediction Run isogd/evaluate.py to get formatted result for evaluation.

##Citation If you use this code in your research, please cite the following two papers:

MFSK feature:
Jun Wan, Guogong Guo, Stan Z. Li, "Explore Efficient Local Features form RGB-D Data for One-shot Learning Gesture Recognition", in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 38, no. 8, pp. 1626-1639, 2016.

IsoGD result: Jun Wan, Yibing Zhao, Shuai Zhou, Isabelle Guyon, Sergio Escalera and Stan Z. Li, "ChaLearn Looking at People RGB-D Isolated and Continuous Datasets for Gesture Recognition", CVPR workshop, 2016.

Should you have any question, please contact:
Shuai Zhou: shuaizhou.palm@gmail.com, or
Jun Wan: jun.wan@ia.ac.cn

##License

VLFeat

The kmeans algorithm we used is implemented by VLFeat toolbox.

VLFeat is distributed under the BSD license: Copyright (C) 2007-11, Andrea Vedaldi and Brian Fulkerson Copyright (C) 2012-13, The VLFeat Team All rights reserved.

IsoGD Baseline

Distributed under the Apache License V2.0, please check LICENSE for further information.