Presentation | 2016-05-19 Study of Photo-realistic Face Moving Image Generation from the Text Using the Facial Feature Kazuki Sato, Takashi Nose, Akinori Ito, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In this paper, we propose face moving image synthesis technique based on Hidden Markov model (HMM) using the facial features as a method of face moving image synthesis for the photo-realistic interactive agent implementation. In recent years, it becomes more and more popular for a human to interact with computers using human-like interfaces, such as an interactive agent. In such a situation, realization of more human-like agent is needed. Therefore, focusing on realization of an agent with human-like appearance, we investigate a synthesis method of photo-realistic face moving image synchronized with the speech contents. The conventional method employs the same framework as the HMM speech synthesis for face motion image synthesis; however, there are problems that high cost of data preparation for learning and low quality of the synthesized facial image. In the proposed method, we use Animation Unit (AU) as feature of face image generation, which can be obtained by Kinect. The AU is parameters that express each part of the face. Using AU, we synthesize the face motion image by converting the face feature generated from the HMM face image synthesizer into intensities of pixels using the Deep Neural Network (DNN). In this paper, we investigate the optimal learning conditions for the HMM an |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | face image synthesis / photo-realistic / HMM / DNN / Kinect |
Paper # | IT2016-8,EMM2016-8 |
Date of Issue | 2016-05-12 (IT, EMM) |
Conference Information | |
Committee | IT / EMM |
---|---|
Conference Date | 2016/5/19(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Otaru Economic Center |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | Information Security, Information Theory, Information Hiding, etc. |
Chair | Yasutada Oohama(Univ. of Electro-Comm.) / Akinori Ito(Tohoku Univ.) |
Vice Chair | Tadashi Wadayama(Nagoya Inst. of Tech.) / Masashi Unoki(JAIST) / Masaki Kawamura(Yamaguchi Univ.) |
Secretary | Tadashi Wadayama(Univ. of Electro-Comm.) / Masashi Unoki(Wakayama Univ.) / Masaki Kawamura(Univ. of Electro-Comm.) |
Assistant | Takuya Kusaka(Okayama Univ.) / Motoi Iwata(Osaka Pref. Univ.) / Kazuhiro Kohno(Kansai Univ.) |
Paper Information | |
Registration To | Technical Committee on Information Theory / Technical Committee on Enriched MultiMedia |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Study of Photo-realistic Face Moving Image Generation from the Text Using the Facial Feature |
Sub Title (in English) | |
Keyword(1) | face image synthesis |
Keyword(2) | photo-realistic |
Keyword(3) | HMM |
Keyword(4) | DNN |
Keyword(5) | Kinect |
1st Author's Name | Kazuki Sato |
1st Author's Affiliation | Tohoku University(Tohoku Univ.) |
2nd Author's Name | Takashi Nose |
2nd Author's Affiliation | Tohoku University(Tohoku Univ.) |
3rd Author's Name | Akinori Ito |
3rd Author's Affiliation | Tohoku University(Tohoku Univ.) |
Date | 2016-05-19 |
Paper # | IT2016-8,EMM2016-8 |
Volume (vol) | vol.116 |
Number (no) | IT-33,EMM-34 |
Page | pp.pp.43-48(IT), pp.43-48(EMM), |
#Pages | 6 |
Date of Issue | 2016-05-12 (IT, EMM) |