• Publications
  • Influence
Target-driven visual navigation in indoor scenes using deep reinforcement learning
TLDR
We proposed a deep reinforcement learning (DRL) framework for target-driven visual navigation. Expand
  • 794
  • 73
  • PDF
AI2-THOR: An Interactive 3D Environment for Visual AI
We introduce The House Of inteRactions (THOR), a framework for visual AI research, available at this http URL AI2-THOR consists of near photo-realistic 3D indoor scenes, where AI agents can navigateExpand
  • 264
  • 48
  • PDF
A Diagram is Worth a Dozen Images
TLDR
We study the problem of diagram interpretation, the challenging task of identifying the structure of a diagram and the semantics of its constituents and their relationships. Expand
  • 69
  • 16
  • PDF
Visual Semantic Planning Using Deep Successor Representations
TLDR
A crucial capability of real-world intelligent agents is their ability to plan a sequence of actions to achieve their goals in the visual world. Expand
  • 94
  • 6
  • PDF
RoboTHOR: An Open Simulation-to-Real Embodied AI Platform
TLDR
We introduce RoboTHOR to democratize research in interactive and embodied visual AI. Expand
  • 17
  • 3
  • PDF
Two Body Problem: Collaborative Visual Task Completion
TLDR
We study the problem of learning to collaborate directly from pixels in AI2-THOR and demonstrate the benefits of explicit and implicit modes of communication to perform visual tasks. Expand
  • 17
  • PDF
A Cordial Sync: Going Beyond Marginal Policies for Multi-Agent Embodied Tasks
TLDR
We introduce the novel task FurnMove in which agents work together to move a piece of furniture through a living room to a goal. Expand
  • 8
  • PDF
Learning Generalizable Visual Representations via Interactive Gameplay
TLDR
We show that embodied adversarial reinforcement learning agents playing cache, a variant of hide-and-seek, in a high fidelity, interactive, environment, learn representations of their observations encoding information such as occlusion, object permanence, free space, and containment; on par with representations learnt by the most popular modern paradigm for visual representation learning which requires large datasets independently labeled for each new task. Expand
  • 7
  • PDF
ManipulaTHOR: A Framework for Visual Object Manipulation
TLDR
We propose a framework for object manipulation built upon the physics-enabled, visually rich AI2-THOR framework and present a new challenge to the Embodied AI community known as ArmPointNav. Expand