| Presentation | 2012-03-08 Word lip reading from scenes of speaker's utterance profile based on mouth-shape-code approach Shinsuke OKITA, Yuki SATO, Yuki SUGATA, Takuro TASAKA, Nozomu HAMADA, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In this paper, we apply mouth-shape-approach to Japanese speaker's utterance profile for lip reading.The novel point is to propose automatic detection of consonant-key-frames. To detect the consonant-key-frames by time series of profile feature vector which is defined the difference value of distance of lips and projection length of lower lip. This approach provides an extension of mouth-shape-code time series. The mouth-shape recognition of key-frames is conducted by five profile shape features; the height of upper lip and lower lip, the projection length of upper and lower lip points, and the angle of lips. We apply DP-matching to the recognized word code string of key-frames and a candidate word code string, then search the nearest word as the result. Recognition experiments using two sets of target 27 words commonly used in dairy conversation, and adding 10 pairs of similar words to them are conducted. The proposed method attained 90.4%, and 86.7% for these word set respectively. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | lip reading / mouth-shape-code / key-frame / profile / image processing |
| Paper # | CAS2011-112,SIP2011-132,CS2011-104 |
| Date of Issue |
| Conference Information | |
| Committee | CAS |
|---|---|
| Conference Date | 2012/3/1(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 | Circuits and Systems (CAS) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Word lip reading from scenes of speaker's utterance profile based on mouth-shape-code approach |
| Sub Title (in English) | |
| Keyword(1) | lip reading |
| Keyword(2) | mouth-shape-code |
| Keyword(3) | key-frame |
| Keyword(4) | profile |
| Keyword(5) | image processing |
| 1st Author's Name | Shinsuke OKITA |
| 1st Author's Affiliation | Department of System Design Engineering, Faculty of Science and Technology, Keio University() |
| 2nd Author's Name | Yuki SATO |
| 2nd Author's Affiliation | Signal processing Lab, School of Integrated Design Engineering, Keio University |
| 3rd Author's Name | Yuki SUGATA |
| 3rd Author's Affiliation | Signal processing Lab, School of Integrated Design Engineering, Keio University |
| 4th Author's Name | Takuro TASAKA |
| 4th Author's Affiliation | Signal processing Lab, School of Integrated Design Engineering, Keio University |
| 5th Author's Name | Nozomu HAMADA |
| 5th Author's Affiliation | Department of System Design Engineering, Faculty of Science and Technology, Keio University |
| Date | 2012-03-08 |
| Paper # | CAS2011-112,SIP2011-132,CS2011-104 |
| Volume (vol) | vol.111 |
| Number (no) | 465 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |