AI2-THOR: An Interactive 3D Environment for Visual AI

by   Eric Kolve, et al.

We introduce The House Of inteRactions (THOR), a framework for visual AI research, available at AI2-THOR consists of near photo-realistic 3D indoor scenes, where AI agents can navigate in the scenes and interact with objects to perform tasks. AI2-THOR enables research in many different domains including but not limited to deep reinforcement learning, imitation learning, learning by interaction, planning, visual question answering, unsupervised representation learning, object detection and segmentation, and learning models of cognition. The goal of AI2-THOR is to facilitate building visually intelligent models and push the research forward in this domain.


page 1

page 2


IQA: Visual Question Answering in Interactive Environments

We introduce Interactive Question Answering (IQA), the task of answering...

iGibson, a Simulation Environment for Interactive Tasks in Large Realistic Scenes

We present iGibson, a novel simulation environment to develop robotic so...

VRKitchen: an Interactive 3D Virtual Environment for Task-oriented Learning

One of the main challenges of advancing task-oriented learning such as v...

IFR-Explore: Learning Inter-object Functional Relationships in 3D Indoor Scenes

Building embodied intelligent agents that can interact with 3D indoor en...

AllenAct: A Framework for Embodied AI Research

The domain of Embodied AI, in which agents learn to complete tasks throu...

The Replica Dataset: A Digital Replica of Indoor Spaces

We introduce Replica, a dataset of 18 highly photo-realistic 3D indoor s...

Unraveling the ARC Puzzle: Mimicking Human Solutions with Object-Centric Decision Transformer

In the pursuit of artificial general intelligence (AGI), we tackle Abstr...

Code Repositories


An open-source platform for Visual AI.

view repo

Please sign up or login with your details

Forgot password? Click here to reset