Robust 3D hand pose estimation from single depth images using multi-view CNNs
Articulated hand pose estimation is one of core technologies in human-computer interaction. Despite the recent progress, most existing methods still cannot achieve satisfactory performance, partly due to the difficulty of the embedded high-dimensional nonlinear regression problem. Most existing data...
| Main Authors: | , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2020
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/140529 |
| _version_ | 1826112525601603584 |
|---|---|
| author | Ge, Liuhao Liang, Hui Yuan, Junsong Thalmann, Daniel |
| author2 | Interdisciplinary Graduate School (IGS) |
| author_facet | Interdisciplinary Graduate School (IGS) Ge, Liuhao Liang, Hui Yuan, Junsong Thalmann, Daniel |
| author_sort | Ge, Liuhao |
| collection | NTU |
| description | Articulated hand pose estimation is one of core technologies in human-computer interaction. Despite the recent progress, most existing methods still cannot achieve satisfactory performance, partly due to the difficulty of the embedded high-dimensional nonlinear regression problem. Most existing data-driven methods directly regress 3D hand pose from 2D depth image, which cannot fully utilize the depth information. In this paper, we propose a novel multi-view convolutional neural network (CNN)-based approach for 3D hand pose estimation. To better exploit 3D information in the depth image, we project the point cloud generated from the query depth image onto multiple views of two projection settings and integrate them for more robust estimation. Multi-view CNNs are trained to learn the mapping from projected images to heat-maps, which reflect probability distributions of joints on each view. These multi-view heat-maps are then fused to estimate the optimal 3D hand pose with learned pose priors, and the unreliable information in multi-view heat-maps is suppressed using a view selection method. Experimental results show that the proposed method is superior to the state-of-the-art methods on two challenging data sets. Furthermore, a cross-data set experiment also validates that our proposed approach has good generalization ability. |
| first_indexed | 2024-10-01T03:08:29Z |
| format | Journal Article |
| id | ntu-10356/140529 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T03:08:29Z |
| publishDate | 2020 |
| record_format | dspace |
| spelling | ntu-10356/1405292020-05-30T06:15:27Z Robust 3D hand pose estimation from single depth images using multi-view CNNs Ge, Liuhao Liang, Hui Yuan, Junsong Thalmann, Daniel Interdisciplinary Graduate School (IGS) Institute for Media Innovation (IMI) Engineering::Computer science and engineering Three-dimensional Displays Heating Systems Articulated hand pose estimation is one of core technologies in human-computer interaction. Despite the recent progress, most existing methods still cannot achieve satisfactory performance, partly due to the difficulty of the embedded high-dimensional nonlinear regression problem. Most existing data-driven methods directly regress 3D hand pose from 2D depth image, which cannot fully utilize the depth information. In this paper, we propose a novel multi-view convolutional neural network (CNN)-based approach for 3D hand pose estimation. To better exploit 3D information in the depth image, we project the point cloud generated from the query depth image onto multiple views of two projection settings and integrate them for more robust estimation. Multi-view CNNs are trained to learn the mapping from projected images to heat-maps, which reflect probability distributions of joints on each view. These multi-view heat-maps are then fused to estimate the optimal 3D hand pose with learned pose priors, and the unreliable information in multi-view heat-maps is suppressed using a view selection method. Experimental results show that the proposed method is superior to the state-of-the-art methods on two challenging data sets. Furthermore, a cross-data set experiment also validates that our proposed approach has good generalization ability. NRF (Natl Research Foundation, S’pore) MOE (Min. of Education, S’pore) Accepted version 2020-05-30T06:11:41Z 2020-05-30T06:11:41Z 2018 Journal Article Ge, L., Liang, H., Yuan, J., & Thalmann, D. (2018). Robust 3D hand pose estimation from single depth images using multi-view CNNs. IEEE Transactions on Image Processing, 27(9), 4422-4436. doi:10.1109/TIP.2018.2834824 1057-7149 https://hdl.handle.net/10356/140529 10.1109/TIP.2018.2834824 9 27 4422 4436 en MOE2015-T2-2-114 IEEE Transactions on Image Processing © 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. The published version is available at: https://doi.org/10.1109/TIP.2018.2834824 application/pdf |
| spellingShingle | Engineering::Computer science and engineering Three-dimensional Displays Heating Systems Ge, Liuhao Liang, Hui Yuan, Junsong Thalmann, Daniel Robust 3D hand pose estimation from single depth images using multi-view CNNs |
| title | Robust 3D hand pose estimation from single depth images using multi-view CNNs |
| title_full | Robust 3D hand pose estimation from single depth images using multi-view CNNs |
| title_fullStr | Robust 3D hand pose estimation from single depth images using multi-view CNNs |
| title_full_unstemmed | Robust 3D hand pose estimation from single depth images using multi-view CNNs |
| title_short | Robust 3D hand pose estimation from single depth images using multi-view CNNs |
| title_sort | robust 3d hand pose estimation from single depth images using multi view cnns |
| topic | Engineering::Computer science and engineering Three-dimensional Displays Heating Systems |
| url | https://hdl.handle.net/10356/140529 |
| work_keys_str_mv | AT geliuhao robust3dhandposeestimationfromsingledepthimagesusingmultiviewcnns AT lianghui robust3dhandposeestimationfromsingledepthimagesusingmultiviewcnns AT yuanjunsong robust3dhandposeestimationfromsingledepthimagesusingmultiviewcnns AT thalmanndaniel robust3dhandposeestimationfromsingledepthimagesusingmultiviewcnns |