CVPR 2024: Language Guided Generation of 3D Embodied AI Environments.

🏘️ Scaling Embodied AI by Procedurally Generating Interactive 3D Houses

🔀 Visual Room Rearrangement

📱👉🏠 Perform conditional procedural generation to generate houses like your own!

🚀 Run AI2-THOR with Google Colab

Implementation of "Safe Deep Learning-Based Global Path Planning Using a Fast Collision-Free Path Generator". We present a global path planning method in this project which is based on an LSTM model that predicts safe paths for the desired start and goal points in an environment with polygonal obstacles, using a new loss function (MSE-NER).

Code for the ICRA 2024 Cook2LTL paper on translating free-form cooking recipes to Linear Temporal Logic (LTL) formulae for robot task planning.

3D household task-based dataset created using customised AI2-THOR.

Visual Reaction: Learning to Play Catch with Your Drone

Implementation of "Safe Deep Learning-Based Global Path Planning Using a Fast Collision-Free Path Generator". We present a global path planning method in this project which is based on an LSTM model that predicts safe paths for the desired start and goal points in an environment with polygonal obstacles, using a new loss function (MSE-NER).

Evaluation tasks for ObjectNav models

Visually grounded planning benchmark for multimodal agents

Fork of allenai/procthor. 🏘️ Scaling Embodied AI by Procedurally Generating Interactive 3D Houses

Robothor Environment for RL training, abide `gym` API

AI2-THOR extension: a browser console and Python toolkit to drive any iTHOR scene, run LLM agents (Ollama, English/Persian), YOLO perception, memory/RAG, evaluation and provenance tracing.

Bounded active memory maintenance for long-horizon embodied agents

7 LLM planning methods (CoT, ReAct, Tree-of-Thought…) for embodied AI — benchmarked in AI2-THOR. Supports OpenAI, Anthropic, Ollama.

An Embodied AI agent for Vision-Language Navigation in a 3D simulator. It combines a grounded LLM planner for high-level reasoning with a deterministic navigation layer that guarantees obstacle avoidance and safe movement.