| Presentation | 2007-06-29 SVR-based Robust Pose Estimation of 3D Objects under Local Occlusion and Non-stationary Backgrounds Shingo ANDO, Akira SUZUKI, Takayuki YASUNO, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Several methods for estimating a pose of 3D object from appearance are proposed these days. We proposed SVR-based pose estimation recently. This method can suppress the storage requirement explosion by describing the relationship between an image and a pose as functions. However, the method is unstable under local occlusion and non-stationary backgrounds, because their variations can not be learned by SVR. In this paper, we propose a novel SVR-based method that is robust against local occlusion and non-stationary backgrounds. First, we use a modified gradient feature which is independent of changes in power and reversals of phase. Second, several local areas are created and pose parameters are estimated in each area. The most feasible parameter values are selected by the confidence-weighted median approach. Experiments show that the proposed method is more stable than the traditional method under several different environments. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | pose estimation / support vector regression(SVR) / principle component analysis / weighted median |
| Paper # | DE2007-14,PRMU2007-40 |
| Date of Issue |
| Conference Information | |
| Committee | DE |
|---|---|
| Conference Date | 2007/6/21(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 | Data Engineering (DE) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | SVR-based Robust Pose Estimation of 3D Objects under Local Occlusion and Non-stationary Backgrounds |
| Sub Title (in English) | |
| Keyword(1) | pose estimation |
| Keyword(2) | support vector regression(SVR) |
| Keyword(3) | principle component analysis |
| Keyword(4) | weighted median |
| 1st Author's Name | Shingo ANDO |
| 1st Author's Affiliation | NTT Cyber Space Laboratories, NTT Corporation() |
| 2nd Author's Name | Akira SUZUKI |
| 2nd Author's Affiliation | NTT Cyber Space Laboratories, NTT Corporation |
| 3rd Author's Name | Takayuki YASUNO |
| 3rd Author's Affiliation | NTT Cyber Space Laboratories, NTT Corporation:Intellectual Property Center, NTT Corporation |
| Date | 2007-06-29 |
| Paper # | DE2007-14,PRMU2007-40 |
| Volume (vol) | vol.107 |
| Number (no) | 114 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |