The Maryland Robotics Center Research Symposium May 2022

MRC Main PagePresentation Recordings
Jana Kosecka Headshot

Jana Kosecka, Professor, George Mason University

 

Vikas Shivashankar Headshot

Vikas Shivashankar, Senior Research Scientist, Amazon Robotics

 

 

Visual Representations for Navigation and Object Discovery

 

Deliberate navigation in previously unseen environments and detection of novel objects instances are some of the key functionalities of intelligent agents engaged in fetch and delivery tasks. While data-driven deep learning approaches fueled rapid progress in object category recognition and semantic segmentation by exploiting large amounts of labelled data, extending this learning paradigm to robotic setting comes with challenges.

To overcome the need for large amount of labeled data for training object instance detectors we use active self-supervision provided by a robot traversing an environment. The knowledge of ego-motion enables the agent to effectively associate multiple object hypotheses, which serve as training data for learning novel object embeddings from unlabelled data. The object detectors trained in this manner achieve higher mAP compared to off-the-shelf detectors trained on this limited data. 

I will describe an approach towards semantic target driven navigation, which entails finding a way through a complex environment to a target object. The proposed approach learns navigation policies on top of representations that capture spatial layout and semantic contextual cues. The choice of this representation exploits models trained on large standard vision datasets, enables better generalization and joint use of simulated environments and real images for effective training of navigation policies. 

 

Jana Kosecka is Professor at the Department of Computer Science, George Mason University. She obtained PhD in Computer Science from University of Pennsylvania. Following her PhD, she was a postdoctoral fellow at the EECS Department at University of California, Berkeley. She is the recipient of David Marr's prize  and received the National Science Foundation CAREER Award. Jana is a chair of IEEE technical Committee of Robot Perception, Associate Editor of IEEE Robotics and Automation Letters and International Journal of Computer Vision, former editor of IEEE Transactions on Pattern Analysis and Machine Intelligence. She held visiting positions at Stanford University, Google and Nokia Research. She  is a co-author of a monograph titled Invitation to 3D vision: From Images to Geometric Models. Her general research interests are in Computer Vision and Robotics. In particular she is interested 'seeing' systems engaged in autonomous tasks, acquisition of static and dynamic models of environments by means of visual sensing and human-computer interaction.

 

Coordinating the Actions of Amazon's Robotic Fleet

 

Every day, Amazon is able to quickly pick, pack, and ship millions of items to customers from a network of fulfillment centers all over the globe. This would not be possible without leveraging cutting-edge advances in planning, scheduling and robotics. In this talk, I will provide an overview of some of these techniques, as well as describe some open research problems we are tackling to further push the envelope for robotics and automation in Amazon's network.

 

Vikas Shivashankar is a Senior Research Scientist at Amazon Robotics. He received his M.S. and PhD. in Computer Science from the University of Maryland, and his B.Tech in Computer Science from IIT Madras. His research is in planning and optimization, with an emphasis on applying these techniques to robotic systems. He has published papers at several top venues including JAIR, AAAI, IJCAI, AAMAS, and won the Outstanding Student Paper award at ICAPS 2014.

Agenda Items

Time

Introduction

8:30-8:50 am

Keynote 1: Jana Kosecka

8:50-9:30 am

Technical Session 1:

9:30-10:10 am

Coffee Break/Poster Session

10:10-10:30 am

Technical Session 2:

10:30-11:10 am

Poster Session

11:10-12:00 pm

Lunch Break

12:00-1:20 pm

Keynote 2: Vikas Shivashankar

1:30-2:10 pm

Technical Session 3:

2:10-3:10 pm

Coffee Break

3:10-3:30 pm

Technical Session 4:

3:30-4:10 pm

Spotlight Session (speakers and video links below) 4:10-4:40 pm

Closing

4:40 pm

Ryan D. Sochol

Associate Professor
301-405-6928 | rsochol@umd.edu
Profile

Pratap Tokekar

Associate Professor
| tokekar@umd.edu
Profile

John S. Baras

Distinguished University Professor
301.405.6606 | baras@umd.edu
Profile

Yancy Diaz-Mercado

Assistant Professor
301-405-6506 | yancy@umd.edu
Profile

Nikhil Chopra

Professor
301.405.7011 | nchopra@umd.edu
Profile

Michael Otte

Assistant Professor
301-405-2723 | otte@umd.edu
Profile

Cecilia Huertas Cerdeira

Assistant Professor
301-405-6640 | huertas@umd.edu
Profile

Yiannis Aloimonos

Professor
301.405.1743 | jyaloimo@umd.edu
Profile

Presenting Authors in bold.

  • HyRiSo: A Hybrid Rigid-Soft Robot for Task Space Manipulation.  Lasitha Weerakoon and Nikhil Chopra.
  • Interactive Multi-Robot Aerial Cinematography Through Hemispherical Manifold Coverage.  Xiaotian Xu, Guangyao Shi, Pratap Tokekar, and Yancy Diaz-Mercado.
  • VTOL Vehicle Design for VFS 2022 DBVF.  Tom Brosh, Paul Zaidins, Qingwen Wei, Dogyu Jun, Animesh Shastry, Wei Cui, and Derek Paley.
  • Development of Fast-Find Strategies for 2021 First Responder UAS Triple Challenge.  Tom Brosh, Paul Zaidins, Qingwen Wei, Dogyu Jun, Animesh Shastry, Wei Cui, and Derek Paley.
  • Data Driven Metareasoning for Defending a Perimeter Against Cooperative Intrusion.  Prannoy Namala, Arjun Vaidya, and Jeffrey Herrmann.
  • Development of Multi-phase Searching Strategies for Multi-agent Systems.  Wei Cui and Derek Paley.
  • PiP-X: Funnel-based Online Feedback Motion Planning/Replanning in Dynamic Environments.  Mohamed Khalid M Jaffar and Michael Otte.
  • Metareasoning Approaches to Thermal Management During Image Processing.  Michael Dawson and Jeffrey Herrmann.
  • Vision-aided Underwater Localization Using Invariant Extended Kalman Filtering.  Tianchen Liu, Miao Yu, and Nikhil Chopra.
  • Simulation of Intermittent Gliding for a Flapping-Wing UAV.  Dior Seck, Lena Johnson, Derek Paley, and Hugh Bruck.
  • Soft Robot Using Fluidic Artificial Muscles for Martian Ice Mining.  Jacek Garbulinski, Frank Cianciarulo, and Norman M. Wereley.
  • Risk-Aware UAV-UGV Rendezvous with Chance-Constrained Markov Decision Process.  Guangyao Shi, Nare Karapetyan, Ahmad Bilal Asghar, Jean-Paul Reddinger, James Dotterweich, James Humann, and Pratap Tokekar.
  • Dynamics and Control of a Buoyancy-Driven Underwater Vehicle for Estimating and Tracking the Scattering Layer.  Rachel J. Suitor, Eric Berkenpas, Charles M. Shepard, Kyler Abernathy, and Derek A. Paley.
  • Kilobot Swarms: Identifying the Effect of Disruptive Behavior.  Chris Martinez Reina, Dalan Loudermilk, and Michael Otte.
  • Sim-to-Real Strategy for Spatially Aware Robot Navigation in Uneven Outdoor Environments.  Kasun Weerakoon, Adarsh Jagan Sathyamoorthy, and Dinesh Manocha.
  • Occupancy Map Prediction for Improved Indoor Robot Navigation.  Vishnu Dutt Sharma, Jingxi Chen, Abhinav Shrivastava, and Pratap Tokekar.
  • Crowd-Steer: Navigation in Crowded and Complex Scenarios.  Jing Liang, Utsav Patel, Adarsh Jagan Sathyamoorthy, and Dinesh Manocha.
  • Dealing with Sparse Rewards in Continuous Control Robotics via Heavy-Tailed Policies.  Souradip Chakraborty, Amrit Singh Bedi, Alec Koppel, Pratap Tokekar, and Dinesh Manocha.
  • ReZoom: Self-driving Autonomous E-scooters.  Srijal Shekhar Poojari, Siddharth Avinash Telang, Andrew Giorgi, and Derek Paley.
  • Fast Active Monocular Distance Estimation from Time-to-Contact.  Levi Burner, Nitin J. Sanket, Cornelia Fermüller, and Yiannis Aloimonos.
Maryland Robotics Center Unmanned Aircraft Systems Test Site
Airgility Kick Robotics

Event begins on May 31st, 2022 starting at 8:30 AM

The symposium presentations will take place in 1101 A. James Clark Hall Forum. The coffee breaks, poster session, and technology showcase will take place across the street at the Kim Engineering Building lobby.

The A. James Clark Hall is building 429 (grid F8) in the campus map.

Google Maps Directions: 

Yancy Diaz-Mercado
Pratap Tokekar
Cornelia Fermüller
Dinesh Manocha
Ania Picard


Top