Paper Details

We introduce a method for model learning and control of non-linear dynamical systems from raw pixelimages. It consists of a deep generative model, belonging to the family of variational autoencoders, that learns togenerate image trajectories from a latent space in which the dynamics is constrained to be locally linear. Our model isderived directly from an optimal control formulation in latent space, supports long-term prediction of image sequencesand exhibits strong performance on a variety of complex control problems.For capturing the information of non-linearobject’s behavior, we need to use high-dimensional data. Processing the high-dimensional data is expensive and notfeasible. So, in this model, first Auto-encoder is used for dimensionality reduction, and after prediction method(transition mapping) is used, and the imagereconstructed. We demonstrate that our model enables learning goodpredictive models of dynamical systems from pixel information only.

Amit Chahar, Anil Kumar, Rohit Rajput, Venkata Ramesh L, Prof. Sunil Dhore, “Learning Non-linear Dynamical Systems From Raw Images”, International Journal of Advance Engineering and Research Development (IJAERD), Vol. 3, Issue 4, pp. 116-119, April 2016.