Abstract
A deep encoder-decoder network was previously proposed for learning a mapping from raw images to dynamic movement primitives in order to enable a robot to draw sketches of numeric digits when shown images of same. In this paper, the network architecture, which was previously constructed entirely with fully-connected linear layers, is modified to include convolutional layers in order to improve the image encoder component and make the network more robust to noise. The convolutional layers are pre-trained as part of an MNIST digit classifier and adapted for use in the encoder-decoder network, before the network is trained using a dataset composed of digit images and corresponding writing trajectories. This architecture was tested on several challenging noisy digit datasets and the use of convolutional layers is shown to provide a robust improvement in results.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Ali, O.: Robotic calligraphy: learning from character images. Ph.D. thesis (2015)
Badrinarayanan, V., Kendall, A., Cipolla, R.: SegNet: a deep convolutional encoder-decoder architecture for image segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 39(12), 2481–2495 (2017)
Basu, S., Karki, M., Ganguly, S., DiBiano, R., Mukhopadhyay, S., Gayaka, S., Kannan, R., Nemani, R.: Learning sparse feature representations using probabilistic quadtrees and deep belief nets. Neural Process. Lett. 45(3), 855–867 (2017)
Chen, N., Bayer, J., Urban, S., van der Smagt, P.: Efficient movement representation by embedding dynamic movement primitives in deep autoencoders. In: 2015 IEEE-RAS 15th International Conference on Humanoid Robots (Humanoids), pp. 434–440 (2015)
Chen, N., Karl, M., van der Smagt, P.: Dynamic movement primitives in latent space of time-dependent variational autoencoders. In: 2016 IEEE-RAS 16th International Conference on Humanoid Robots (Humanoids), pp. 629–636 (2016)
Ijspeert, A.J., Nakanishi, J., Hoffmann, H., Pastor, P., Schaal, S.: Dynamical movement primitives: learning attractor models for motor behaviors. Neural Comput. 25(2), 328–373 (2013)
Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. In: 3rd International Conference for Learning Representations (ICLR), San Diego (2015)
Krüger, N., Geib, C., Piater, J., Petrick, R., Steedman, M., Wörgötter, F., Ude, A., Asfour, T., Kraft, D., Omrčen, D., Agostini, A., Dillmann, R.: Object-action complexes: grounded abstractions of sensory-motor processes. Robot. Auton. Syst. 59(10), 740–757 (2011)
Pahič, R., Gams, A., Ude, A., Morimoto, J.: Deep encoder-decoder networks for mapping raw images to dynamic movement primitives. In: 2018 IEEE International Conference on Robotics and Automation (ICRA), Brisbane, Australia, pp. 5863–5868 (2018)
Paszke, A., Gross, S., Chintala, S., Chanan, G., Yang, E., DeVito, Z., Lin, Z., Desmaison, A., Antiga, L., Lerer, A.: Automatic differentiation in PyTorch. In: NIPS 2017 Autodiff Workshop (2017)
Pervez, A., Mao, Y., Lee, D.: Learning deep movement primitives using convolutional neural networks. In: 2017 IEEE-RAS 17th International Conference on Humanoid Robotics (Humanoids), pp. 191–197 (2017)
Ude, A., Gams, A., Asfour, T., Morimoto, J.: Task-specific generalization of discrete and periodic dynamic movement primitives. IEEE Trans. Robot. 26(5), 800–815 (2010)
Yang, P.C., Sasaki, K., Suzuki, K., Kase, K., Sugano, S., Ogata, T.: Repeatable folding task by humanoid robot worker using deep learning. IEEE Robot. Autom. Lett. 2(2), 397–403 (2017)
Acknowledgement
This work has received funding from the EU’s Horizon 2020 RIA AUTOWARE (GA no. 723909); the Slovenian Research Agency under GA no. J2-7360; JSPS KAKENHI JP16H06565; NEDO; the Commissioned Research of NICT; the NICT Japan Trust (International research cooperation program); and JST-Mirai Program Grant Number JPMJMI18B8, Japan. The authors also wish to thank Marcel Salmič for his significant contribution to the PyTorch network implementations.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Ridge, B., Pahič, R., Ude, A., Morimoto, J. (2020). Convolutional Encoder-Decoder Networks for Robust Image-to-Motion Prediction. In: Berns, K., Görges, D. (eds) Advances in Service and Industrial Robotics. RAAD 2019. Advances in Intelligent Systems and Computing, vol 980. Springer, Cham. https://doi.org/10.1007/978-3-030-19648-6_59
Download citation
DOI: https://doi.org/10.1007/978-3-030-19648-6_59
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-19647-9
Online ISBN: 978-3-030-19648-6
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)