CoRL 2020, Spotlight Talk 448: DeepMPCVS: Deep Model Predictive Control for Visual Servoing

CoRL 2020

CoRL 2020, Spotlight Talk 448: DeepMPCVS: Deep Model Predictive Control for Visual Servoing

Dec 16, 2020
|
24 views
Details
"**DeepMPCVS: Deep Model Predictive Control for Visual Servoing** Pushkal Katara (Robotics Research Center, IIITH)*; Harish Y V S (IIIT HYDERABAD); Harit Pandya (University of Lincoln); Abhinav Gupta (International Institute of Information Technology (IIIT), Hyderabad); AadilMehdi Sanchawala (International Institute of Information Technology, Hyderabad); Gourav Kumar (TCS Innovation labs Kolkata); Brojeshwar Bhowmick (Tata Consultancy Services); Madhava Krishna (IIIT-Hyderabad) Publication: http://corlconf.github.io/paper_448/ **Abstract** The simplicity of the visual servoing approach makes it an attractive option for tasks dealing with vision-based control of robots in many real-world applications. However, attaining precise alignment for unseen environments pose a challenge to existing visual servoing approaches. While classical approaches assume a perfect world, the recent data-driven approaches face issues when generalizing to novel environments. In this paper, we aim to combine the best of both worlds. We present a deep model predictive visual servoing framework that can achieve precise alignment with optimal trajectories and can generalize to novel environments. Our framework consists of a deep network for optical flow predictions, which are used along with a predictive model to forecast future optical flow. For generating an optimal set of velocities we present a control network that can be trained on-the-fly without any supervision. Through extensive simulations on photo-realistic indoor settings of the popular Habitat framework, we show significant performance gain due to the proposed formulation vis-a-vis recent state of the art methods. Specifically, we show vastly improved performance in trajectory length and faster convergence over recent approaches."

Comments
loading...