ChaLearn LAP IsoGD and ConGD datasets

We have built two large-scale gesture datasets: ChaLearn LAP Isolated Gesture Dataset (IsoGD) and ChaLearn LAP Continous Gesture Dataset (ConGD). The focus of the challenges is "large-scale" learning and "user independent" gesture recogniton form RGB or RGB-D videos.

Both dataset was created from the CGD 2011 dataset.

 

1. Isolated Gesture Recognition Challenge

Database Infomation and Format

This database includes 47933 RGB-D gesture videos (about 9G). Each RGB-D video represents one gesture only, and there are 249 gestures labels performed by 21 different individuals.

The database has been divided to three sub-datasets for the convenience of using, and these three subsets are mutually exclusive.

Sets# of Labels# of Gestures# of RGB Vidoes# of Depth Vidoes# of PerformersLabel Provided
Training24935878358783587817Yes
Validation2495784578457842No
Testing2496271627162712No

Metric Evaluation: Recognition Rate

Main Task:

1) isolate gesture recognition using RGB and depth videos

2)  Large scale: 47749 gestures and 249 labels

3) User Independent:  the uses in training set will not disappear in testing and validation set.

Baseline Method:  BOW+MFSK+SVM and BOW+(MFSK+deep ID)+SVM

2. Continuous Gesture Recognition Challenge

Database Infomation and Format

This database includes 47933 RGB-D gestures in 22535 RGB-D gesture videos (about 4G). Each RGB-D video may represent one or more gestures, and there are 249 gestures labels performed by 21 different individuals.

The database has been divided to three sub-datasets for the convenience of using, and these three subsets are mutually exclusive.

Sets# of Labels# of Gestures# of RGB Vidoes# of Depth Vidoes# of PerformersLabel ProvidedTemporal Segmentation Provided
Training24930442143141431417YesYes
Validation2498889417941792NoNo
Testing2498602404240422NoNo

Metric Evaluation: Jaccard Index

Main Task:

1) gesture spotting and recognition from continuous RGB and depth videos 

2)  Large scale:  47933 gestures in  22535 RGB-Depth videos, 249 labels

3) User Independent:  the uses in training set will not disappear in testing and validation set.

Baseline Method:   BOW+MFSK+SVM+sliding window and BOW+(MFSK+deep ID)+SVM+sliding window


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

Apply both datasets: 

The ChaLearn LAP IsoGD dataset

The ChaLearn LAP ConGD dataset

Comments