Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor | Faculty of Engineering | en_US |
| dc.creator | Li, Junxian | - |
| dc.identifier.uri | https://theses.lib.polyu.edu.hk/handle/200/10102 | - |
| dc.language | English | en_US |
| dc.publisher | Hong Kong Polytechnic University | - |
| dc.rights | All rights reserved | en_US |
| dc.title | Accumulated effective optical flow feature for micro-expression recognition | en_US |
| dcterms.abstract | Micro-expressions are very brief facial expressions which have short duration, and low intensity and arise at small, partial and irregular region of a face. Compared to the basic facial expression recognition, micro-expression is more complex and harder to be detected and recognized by human's eyes and computer-vision methods. This dissertation proposes an Accumulated Effective Optical flow (AEOF) feature for micro-expression recognition. Firstly, the integrated optical flow fields between every two adjacent frames are computed. Then, 68 facial landmarks of a face subject in a video sequence are detected and located. Based on the landmarks, eleven regions of interest (ROIs) are partitioned according to the Facial Action Coding System (FACS). At the same time, the onset frames and apex frames of all micro-expression video samples are spotted, based on the mean optical-flow magnitude (MOFM) feature in the eleven ROIs. Next, the AEOF features, which have a dimension of 11×8=88, are extracted from onset frame to apex frame. Finally, Support Vector Machine (SVM) classifiers with the AEOF features are applied for micro-expression recognition. Evaluating on the spontaneous micro-expression database CASME II, experiment results show the AEOF can recognize micro-expressions accurately, and achieve an improvement in comparison with an optical-flow baseline feature, called Histogram of Optical-Flow (HOOF), and a state-of-the-art method, Main Directional Mean Optical-Flow (MDMO). | en_US |
| dcterms.extent | iv, 33 pages : color illustrations | en_US |
| dcterms.isPartOf | PolyU Electronic Theses | en_US |
| dcterms.issued | 2019 | en_US |
| dcterms.educationalLevel | M.Sc. | en_US |
| dcterms.educationalLevel | All Master | en_US |
| dcterms.LCSH | Hong Kong Polytechnic University -- Dissertations | en_US |
| dcterms.LCSH | Human face recognition (Computer science) | en_US |
| dcterms.LCSH | Image analysis -- Data processing | en_US |
| dcterms.LCSH | Image processing -- Digital techniques | en_US |
| dcterms.accessRights | restricted access | en_US |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 991022270857203411.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 772.37 kB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/10102