| Presentation | 2023-03-02 Blink detection from one-dimensional face signal by using convolutional sparse dictionary learning Souichiro Maruyama, Makoto Nakashizuka, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In this report, a blink detection method from average intensities of whole facial videos using convolutional dictionary learning is presented. Recently, blink detection has been used to measure health and psychological states. Existing blink detection methods extract feature points of a face and detect blinks from their geometric changes. However, this methodrequires high-resolution images because the feature points of the eyes must be accurately captured. Therefore, this method is considered inappropriate from the viewpoint of personal information protection because it can identify individuals. In this report, we aim to detect blink from low-resolution images, which is impossible to be used for personal identification. Specifically, a one-dimensional temporal signal is obtained from the average intensity over the entire face at each frame, and blinks are detected from this one-dimensional signal. This problem is a signal separation problem from a mixture of signals. The blink separation method In our detection, the variations of the intensity caused by the blink is extracted as the dictionaries of the sparse coding, prior to separation. The separation is performed by the sparse coding with the dictionary that is trained. In experiments, the proposed method is applied to detect blinks, and the detection accuracy is evaluated based on the fit rate and reproduction rate. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Blink detection / blind source separation / sparse representation / dictionary learning |
| Paper # | SIS2022-40 |
| Date of Issue | 2023-02-23 (SIS) |
| Conference Information | |
| Committee | SIS |
|---|---|
| Conference Date | 2023/3/2(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Chiba Institute of Technology |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Tomoaki Kimura(Kanagawa Inst. of Tech.) |
| Vice Chair | Naoto Sasaoka(Tottori Univ.) / Hakaru Tamukoh(Kyushu Inst. of Tech.) |
| Secretary | Naoto Sasaoka(NTT) / Hakaru Tamukoh(Kansai Univ.) |
| Assistant | Yoshiaki Makabe(Kanagawa Inst. of Tech.) / Yosuke Sugiura(Saitama Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Smart Info-Media Systems |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Blink detection from one-dimensional face signal by using convolutional sparse dictionary learning |
| Sub Title (in English) | |
| Keyword(1) | Blink detection |
| Keyword(2) | blind source separation |
| Keyword(3) | sparse representation |
| Keyword(4) | dictionary learning |
| 1st Author's Name | Souichiro Maruyama |
| 1st Author's Affiliation | Chiba Institute of Technology(CIT) |
| 2nd Author's Name | Makoto Nakashizuka |
| 2nd Author's Affiliation | Chiba Institute of Technology(CIT) |
| Date | 2023-03-02 |
| Paper # | SIS2022-40 |
| Volume (vol) | vol.122 |
| Number (no) | SIS-410 |
| Page | pp.pp.1-4(SIS), |
| #Pages | 4 |
| Date of Issue | 2023-02-23 (SIS) |