search

UMD     This Site





The agent learns to execute instructions by synthesizing and reaching visual goals from a learned latent variable model. (Figure 1 from the paper)

The agent learns to execute instructions by synthesizing and reaching visual goals from a learned latent variable model. (Figure 1 from the paper)

 

Building robotic agents that can follow instructions in physical environments is one of the most enduring challenges in robotics and artificial intelligence, dating back to the 1970s. Traditionally, methods for accomplishing this assume the agent has prior linguistic, perceptual, and procedural knowledge. However, recently deep reinforcement learning (RL) has emerged as a promising framework for instruction-following tasks. In this method, policies could be learned end-to-end, typically by training neural networks to map joint representations of observations and instructions directly to actions.

To design agents that exhibit intelligence and learn through interaction, roboticists will need to incorporate cognitive capabilities like imagination, one of the hallmarks of human-level intelligence.

Following Instructions by Imagining and Reaching Visual Goals, a new paper by John Kanu, Eadom Dessalene, Xiaomin Lin, ISR-affiliated Associate Research Scientist Cornelia Fermüller (UMIACS), and ISR-affiliated Professor Yiannis Aloimonos (CS/UMIACS) addresses the need to incorporate explicit spatial reasoning to accomplish temporally extended tasks with minimal manual engineering.

The researchers present a novel framework for learning to perform temporally extended tasks using spatial reasoning in the RL framework, by sequentially imagining visual goals and choosing appropriate actions to fulfill imagined goals. The framework, called Instruction-Following with Imagined Goals (IFIG), integrates a mechanism for sequentially synthesizing visual goal states and methods for self-supervised state representation learning and visual goal reaching. IFIG requires minimal manual engineering, operates on raw pixel images, and assumes no prior linguistic or perceptual knowledge. It learns via intrinsic motivation and a single extrinsic reward signal measuring task completion.

The authors validate their method in two environments with a robot arm in a simulated interactive 3D environment, and find IFIG is able to decompose object-arrangement tasks into a series of visual states describing sub-goal arrangements. Their method outperforms two flat architectures with raw-pixel and ground-truth states, and a hierarchical architecture with ground-truth states on object arrangement tasks.

Without any labelling of ground-truth state, IFIG outperforms methods that have access to ground-truth state, as well as methods without goal-setting or explicit spatial reasoning. Through visualization of goal reconstructions, the authors show their model has learned instruction-conditional state transformations, suggesting an ability to perform task-directed spatial reasoning.



Related Articles:
A learning algorithm for training robots' deep neural networks to grasp novel objects
A cooperative control algorithm for robotic search and rescue
Who's walking deceptively? Manocha's team thinks they know.
Do Good Robotics Start-Up Competition coming this fall
Helping robots remember
Realistic simulator improves safety of self-driving vehicles before road testing
Do Good Robotics Symposium to explore technologies that benefit society and the planet
Czech prime minister views AI, VR, AR and computer vision research
NSF funds Shamma, Espy-Wilson for neuromorphic and data-driven speech segregation research
New affiliate faculty Mark Fuge is expert in machine learning and artificial intelligence

February 14, 2020


«Previous Story  

 

 

Current Headlines

In Race With Virus, Researchers Speed Development of Medical Equipment

GAMMA Group's Research on Emotional Modeling and Social Robotics Featured in Forbes

Srivastava wins NSF funding for integrated circuit fabrication security

Protection Collections Abound for Local Health Care Workers

New U.S. Patent: Integrated Onboard Chargers for Plug-In Vehicles

Public health planners: Free resources for emergency health clinics

Planning and learning algorithms developed for refinement acting engine

Clark School Engineers Create Solutions for a Crisis

COVID-19 Decision Making Gets a Big Data Boost

Researchers from Poland report results aided by UMD's AIM Lab

 
 
Back to top  
Home Clark School Home UMD Home