Recognising human actions by analysing negative spaces
The authors propose a novel region-based method to recognise human actions. Other region-based approaches work on silhouette of the human body, which is termed as the positive space according to art theory. In contrast, the authors investigate and analyse regions surrounding the human body, termed a...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2013
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/107336 http://hdl.handle.net/10220/16663 |
| _version_ | 1811691514210811904 |
|---|---|
| author | Rahman, S. A. Leung, Maylor Karhang Cho, Siu-Yeung |
| author2 | School of Computer Engineering |
| author_facet | School of Computer Engineering Rahman, S. A. Leung, Maylor Karhang Cho, Siu-Yeung |
| author_sort | Rahman, S. A. |
| collection | NTU |
| description | The authors propose a novel region-based method to recognise human actions. Other region-based approaches work on silhouette of the human body, which is termed as the positive space according to art theory. In contrast, the authors investigate and analyse regions surrounding the human body, termed as the negative space for human action recognition. This concept takes advantage of the naturally formed negative regions that come with simple shape, simplifying the job for action classification. Negative space is less sensitive to segmentation errors, overcoming some limitations of silhouette-based methods such as leaks or holes in the silhouette caused by background segmentation. Inexpensive semantic-level description can be generated from the negative space that supports fast and accurate action recognition. The proposed system has obtained 100% accuracy on the Weizmann human action dataset and the robust sequence dataset. On KTH dataset the system achieved 94.67% accuracy. Furthermore, 95% accuracy can be achieved even when half of the negative space regions are ignored. This makes our work robust with respect to segmentation errors and distinctive from other approaches. |
| first_indexed | 2024-10-01T06:21:06Z |
| format | Journal Article |
| id | ntu-10356/107336 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T06:21:06Z |
| publishDate | 2013 |
| record_format | dspace |
| spelling | ntu-10356/1073362020-05-28T07:18:45Z Recognising human actions by analysing negative spaces Rahman, S. A. Leung, Maylor Karhang Cho, Siu-Yeung School of Computer Engineering DRNTU::Engineering::Computer science and engineering The authors propose a novel region-based method to recognise human actions. Other region-based approaches work on silhouette of the human body, which is termed as the positive space according to art theory. In contrast, the authors investigate and analyse regions surrounding the human body, termed as the negative space for human action recognition. This concept takes advantage of the naturally formed negative regions that come with simple shape, simplifying the job for action classification. Negative space is less sensitive to segmentation errors, overcoming some limitations of silhouette-based methods such as leaks or holes in the silhouette caused by background segmentation. Inexpensive semantic-level description can be generated from the negative space that supports fast and accurate action recognition. The proposed system has obtained 100% accuracy on the Weizmann human action dataset and the robust sequence dataset. On KTH dataset the system achieved 94.67% accuracy. Furthermore, 95% accuracy can be achieved even when half of the negative space regions are ignored. This makes our work robust with respect to segmentation errors and distinctive from other approaches. 2013-10-21T07:41:05Z 2019-12-06T22:29:04Z 2013-10-21T07:41:05Z 2019-12-06T22:29:04Z 2012 2012 Journal Article Rahman, S.A., Cho, S.-Y., & Leung, M.K. (2012). Recognising human actions by analysing negative spaces. IET computer vision, 6(3), 197-213. 1751-9632 https://hdl.handle.net/10356/107336 http://hdl.handle.net/10220/16663 10.1049/iet-cvi.2011.0185 en IET computer vision |
| spellingShingle | DRNTU::Engineering::Computer science and engineering Rahman, S. A. Leung, Maylor Karhang Cho, Siu-Yeung Recognising human actions by analysing negative spaces |
| title | Recognising human actions by analysing negative spaces |
| title_full | Recognising human actions by analysing negative spaces |
| title_fullStr | Recognising human actions by analysing negative spaces |
| title_full_unstemmed | Recognising human actions by analysing negative spaces |
| title_short | Recognising human actions by analysing negative spaces |
| title_sort | recognising human actions by analysing negative spaces |
| topic | DRNTU::Engineering::Computer science and engineering |
| url | https://hdl.handle.net/10356/107336 http://hdl.handle.net/10220/16663 |
| work_keys_str_mv | AT rahmansa recognisinghumanactionsbyanalysingnegativespaces AT leungmaylorkarhang recognisinghumanactionsbyanalysingnegativespaces AT chosiuyeung recognisinghumanactionsbyanalysingnegativespaces |