| Presentation | 2001/10/18 Facial Expression Recognition Using Vector Matching of Spacial Frequency Components Yoshiko Inada, Yegui Xiao, Masaomi Oda, |
|---|---|
| PDF Download Page | PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In recent years, recognition of facial expressions has been a hot area of research, and a lot of techniques have been proposed. In particular, a method based on the two-dimensional (2-D) face images was proposed, which uses low-frequency coefficients of the 2-D discrete cosine transform (2-D DCT) of the whole image as a feature detector and trains a neural network (NN) as a classifier. The use of the 2-D DCT coefficients unuseful in the recognition enlarges the NN size and therefore increases the cost of NN training. One more problem with this and almost all the previous recognition methods is that they always require the neutral face images in both training and testing phases. In this paper, first, a new recognition method is proposed which uses a mask pattern that is trained to shut out those 2-D DCT coefficients that contribute nothing to the recognition. Next, one more new method is proposed that does not require neutral images in the testing phase, and is therefore referred to as blind facial expression recognition method. Both methods are based on simple vector matching. Experimental results are provided to demonstrate the performance of the two proposed methods. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Facial expressions / 2-D DCT / Vector matching / Mask pattern / Blind facial expression recognition |
| Paper # | DSP2001-103,ICD2001-108,IE2001-87 |
| Date of Issue |
| Conference Information | |
| Committee | DSP |
|---|---|
| Conference Date | 2001/10/18(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Digital Signal Processing (DSP) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Facial Expression Recognition Using Vector Matching of Spacial Frequency Components |
| Sub Title (in English) | |
| Keyword(1) | Facial expressions |
| Keyword(2) | 2-D DCT |
| Keyword(3) | Vector matching |
| Keyword(4) | Mask pattern |
| Keyword(5) | Blind facial expression recognition |
| 1st Author's Name | Yoshiko Inada |
| 1st Author's Affiliation | Hiroshima Prefectural Women's University() |
| 2nd Author's Name | Yegui Xiao |
| 2nd Author's Affiliation | Hiroshima Prefectural Women's University |
| 3rd Author's Name | Masaomi Oda |
| 3rd Author's Affiliation | Ritsumeikan University |
| Date | 2001/10/18 |
| Paper # | DSP2001-103,ICD2001-108,IE2001-87 |
| Volume (vol) | vol.101 |
| Number (no) | 383 |
| Page | pp.pp.- |
| #Pages | 8 |
| Date of Issue | |