Human action recognition using a fast learning fully complex-valued classifier
In this paper, we use optical flow based complex-valued features extracted from video sequences to recognize human actions. The optical flow features between two image planes can be appropriately represented in the Complex plane. Therefore, we argue that motion information that is used to model the...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2013
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/98664 http://hdl.handle.net/10220/13654 |
| _version_ | 1811683105154531328 |
|---|---|
| author | Suresh, Sundaram Venkatesh Babu, R. Savitha, R. |
| author2 | School of Computer Engineering |
| author_facet | School of Computer Engineering Suresh, Sundaram Venkatesh Babu, R. Savitha, R. |
| author_sort | Suresh, Sundaram |
| collection | NTU |
| description | In this paper, we use optical flow based complex-valued features extracted from video sequences to recognize human actions. The optical flow features between two image planes can be appropriately represented in the Complex plane. Therefore, we argue that motion information that is used to model the human actions should be represented as complex-valued features and propose a fast learning fully complex-valued neural classifier to solve the action recognition task. The classifier, termed as, “fast learning fully complex-valued neural (FLFCN) classifier” is a single hidden layer fully complex-valued neural network. The neurons in the hidden layer employ the fully complex-valued activation function of the type of a hyperbolic secant function. The parameters of the hidden layer are chosen randomly and the output weights are estimated as the minimum norm least square solution to a set of linear equations. The results indicate the superior performance of FLFCN classifier in recognizing the actions compared to real-valued support vector machines and other existing results in the literature. Complex valued representation of 2D motion and orthogonal decision boundaries boost the classification performance of FLFCN classifier. |
| first_indexed | 2024-10-01T04:07:26Z |
| format | Journal Article |
| id | ntu-10356/98664 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T04:07:26Z |
| publishDate | 2013 |
| record_format | dspace |
| spelling | ntu-10356/986642020-05-28T07:17:53Z Human action recognition using a fast learning fully complex-valued classifier Suresh, Sundaram Venkatesh Babu, R. Savitha, R. School of Computer Engineering DRNTU::Engineering::Computer science and engineering In this paper, we use optical flow based complex-valued features extracted from video sequences to recognize human actions. The optical flow features between two image planes can be appropriately represented in the Complex plane. Therefore, we argue that motion information that is used to model the human actions should be represented as complex-valued features and propose a fast learning fully complex-valued neural classifier to solve the action recognition task. The classifier, termed as, “fast learning fully complex-valued neural (FLFCN) classifier” is a single hidden layer fully complex-valued neural network. The neurons in the hidden layer employ the fully complex-valued activation function of the type of a hyperbolic secant function. The parameters of the hidden layer are chosen randomly and the output weights are estimated as the minimum norm least square solution to a set of linear equations. The results indicate the superior performance of FLFCN classifier in recognizing the actions compared to real-valued support vector machines and other existing results in the literature. Complex valued representation of 2D motion and orthogonal decision boundaries boost the classification performance of FLFCN classifier. 2013-09-24T07:18:16Z 2019-12-06T19:58:14Z 2013-09-24T07:18:16Z 2019-12-06T19:58:14Z 2012 2012 Journal Article Venkatesh Babu, R., Suresh, S., & Savitha, R. (2012). Human action recognition using a fast learning fully complex-valued classifier. Neurocomputing, 89, 202-212. https://hdl.handle.net/10356/98664 http://hdl.handle.net/10220/13654 10.1016/j.neucom.2012.03.003 en Neurocomputing |
| spellingShingle | DRNTU::Engineering::Computer science and engineering Suresh, Sundaram Venkatesh Babu, R. Savitha, R. Human action recognition using a fast learning fully complex-valued classifier |
| title | Human action recognition using a fast learning fully complex-valued classifier |
| title_full | Human action recognition using a fast learning fully complex-valued classifier |
| title_fullStr | Human action recognition using a fast learning fully complex-valued classifier |
| title_full_unstemmed | Human action recognition using a fast learning fully complex-valued classifier |
| title_short | Human action recognition using a fast learning fully complex-valued classifier |
| title_sort | human action recognition using a fast learning fully complex valued classifier |
| topic | DRNTU::Engineering::Computer science and engineering |
| url | https://hdl.handle.net/10356/98664 http://hdl.handle.net/10220/13654 |
| work_keys_str_mv | AT sureshsundaram humanactionrecognitionusingafastlearningfullycomplexvaluedclassifier AT venkateshbabur humanactionrecognitionusingafastlearningfullycomplexvaluedclassifier AT savithar humanactionrecognitionusingafastlearningfullycomplexvaluedclassifier |