Bidirectional recurrent autoencoder for 3D skeleton motion data refinement. (June 2019)

• Record Type: Journal Article
• Authors: Li, Shujie; Zhou, Yang; Zhu, Haisheng; Xie, Wenjun; Zhao, Yang; Liu, Xiaoping
• Is Part Of: Computers & graphics. Volume 81(2019)
• Page Start: 92
• Page End: 103
• Publication Date: 2019-06
• Subjects: Motion data refinement -- B-LSTM-RNN -- Motion autoencoder -- 3D skeleton motion data -- Joint position; Computer graphics -- Periodicals; 006.6
• DOI: 10.1016/j.cag.2019.03.010
• Languages: English
• ISSNs: 0097-8493

Abstract

Highlights: The proposed BRA can generate motion data that have much less reproduction error compared to the CNN architecture. The proposed BRA does not require postprocessing procedure. The proposed BRA does not require the noise amplitude as a priori knowledge. Graphical abstract: Abstract: In this paper, we propose a novel 3D skeleton human motion data refinement method that is based on a bidirectional recurrent autoencoder (BRA). The BRA has two main characteristics: (1) the motion manifold is extracted by a bidirectional long short-term memory recurrent neural network (B-LSTM-RNN) and (2) apart from statistical information of motion data, kinematic information including smoothness and bone length constrain, are also simultaneously exploited with noisy-clean motion pairs. Using a bidirectional LSTM unit, which is more suitable for time series and can infer information from the data in both time directions, our autoencoder extracts a manifold that can exploit the spatial and temporal relationships between previous and subsequent motion data. As a result, the refined data that are projected by the decoder from the motion manifold have much lower reproduction error. Furthermore, owing to the consideration of kinematic information, our reproduced motion data are of higher visual quality, while preserving positional precision. The proposed method is not action-specific and can handle a wide variety of noise types. The proposed method does not require the noise amplitude, which may be unknown in many scenarios, as a priori knowledge. Extensive experimental results demonstrate that our method outperforms several state-of-the-art methods.