Grasp Type Revisited: A Modern Perspective on A Classical Feature for Vision
Abstract
The grasp type provides crucial information about human action. However, recognizing the grasp type from unconstrained scenes is challenging because of the large variations in appearance, occlusions and geometric distortions.
In this paper, first we present a convolutional neural network to classify functional hand grasp types. Experiments
on a public static scene hand data set validate good performance of the presented method. Then we present two applications utilizing grasp type classification: (a) inference of
human action intention and (b) fine level manipulation action segmentation. Experiments on both tasks demonstrate
the usefulness of grasp type as a cognitive feature for computer vision. This study shows that the grasp type is a powerful symbolic representation for action understanding, and thus opens new avenues for future research.
The Static Image Grasp Type dataset
4672 annotated hand patches are available for training, 660 for testing. Please refer to the readme file for the details.
ZIP file: StaticGrasp1.0
Static image Intention Dataset Intention Data.
Related Publication:
@inproceedings{yang2015grasp,
title={Grasp Type Revisited: A Modern Perspective on A Classical Feature for Vision},
author={Yang, Yezhou and Fermuller, Cornelia and Li, Yi and Aloimonos, Yiannis},
booktitle={Computer Vision and Pattern Recognition (CVPR), 2015 IEEE Conference on},
year={2015},
organization={IEEE}
}
Special thanks to given for sharing their static image hand dataset,
OXFORD HAND DATASET.
Questions? Please contact yzyang "at" cs dot umd dot edu