Multi-layered Gesture Recognition with Kinect
Authors: Feng Jiang, Shengping Zhang, Shen Wu, Yang Gao, Debin Zhao
JMLR 2015 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | This paper proposes a novel multi-layered gesture recognition method with Kinect. ... The proposed method can obtain relatively high performance for one-shot learning gesture recognition on the Cha Learn Gesture Dataset comprising more than 50, 000 gesture sequences recorded with Kinect. ... In this section, extensive experiment results are presented to evaluate the proposed multilayered gesture recognition method. All the experiments are performed in Matlab 7.12.0 on a Dell PC with Duo CPU E8400. The Cha Learn Gesture Dataset (CGD 2011) (Guyon et al., 2014) is used in all experiments, which is designed for one-shot learning. |
| Researcher Affiliation | Academia | School of Computer Science and Technology Harbin Institute of Technology, Harbin 150001, China |
| Pseudocode | Yes | Algorithm 1 Computing weight vector W R for a reference R... Algorithm 2 Computing distance between two unclosed contour point sequences |
| Open Source Code | Yes | The code is available at http://gesture.chalearn.org/data/sample-code |
| Open Datasets | Yes | The Cha Learn Gesture Dataset (CGD 2011) (Guyon et al., 2014) is used in all experiments, which is designed for one-shot learning. |
| Dataset Splits | Yes | The Cha Learn Gesture Dataset (CGD 2011) (Guyon et al., 2014) is used in all experiments, which is designed for one-shot learning. ... An extreme example is the one-shot learning that uses only one training sample per class for training. ... We evaluate the performance of the proposed method with different layers on the development (devel01 devel480) batches of CGD 2011 and Table 3 reports the results. |
| Hardware Specification | Yes | All the experiments are performed in Matlab 7.12.0 on a Dell PC with Duo CPU E8400. |
| Software Dependencies | Yes | All the experiments are performed in Matlab 7.12.0 on a Dell PC with Duo CPU E8400. |
| Experiment Setup | Yes | The parameters used in the proposed method are listed in Table 2. ... For each reference gesture, the final V reconstruction errors are obtained by multiplying the reconstruction errors of the depth data and the gray data. These V reconstruction errors are further clustered by K-means to get two centers. ... we empirically set its value with 0.05 by observing the matching scores between the test sample and each gesture candidates. ... For CHMM and SCHMM, the covariance matrix is a diagonal matrix with all diagonal elements being 0.2. |