In order to provide benchmark data for clothes material/category recognition in free-configurations, we captured a high-quality RGBD clothing dataset using a stereo head system. In our dataset, there are 50 items of clothing from 5 categories:  t-shits, shirts,  sweaters, jeans and towels, of which the material types are: cotton, jaconet, wool, denim, coarse cotton, respectively. Each item of clothing is captured in 21 different random configurations, resulting in a total of 1050 configurations. For each configuration the RGB image, depth map and the segmented mask are provided – both the RGB and depth data are 16 mega-pixels (3264×4928). This is the first high-resolution free-configuration clothing dataset.

Recognition Challenge:

 Method Dataset Feature Classifier Evaluation Criterion Accuracy
 Topological-Shape [1] Robot Head L-S-T-B SVM 5-fold-CV on clothes 83.2%
 Topological-Shape[1] Kinect L-S-T-B SVM 5-fold-CV on clothes 64.2%

 Welcome to share your results with us~


Li SunSimon Rogers, Gerarado Aragon-Camarasa, J. Paul Siebert. "Recognising the Clothing Categories from Free-Configuration using Gaussian-Process-Based Interactive Perception",  ICRA2016.
Li Sun
Gerarado Aragon-Camarasa
Simon Rogers, J. Paul Siebert. "
Single-Shot Clothing Category Recognition in Free-Configurations with Application to Autonomous Clothes Sorting", IROS 2017.

Dataset Download:


Source Code: