Skip to content
@OpenXRIF

OpenXRIF

An open-source embodied AI framework enabling accessible human-robot interaction
README.md

Robots are limited in their ability to interpret natural language and execute tasks in ambiguous environments. We're a small team of engineers and researchers building XRIF (eXtensible Robot Instruction Format) to serve as a framework for social robotic intelligence! Our north star is to democratize a robot-agnostic platform, enabling users to define action spaces, constraints, and guardrails for their human-robot interaction use cases.

Here are some cool papers that inspired us:

Pinned Loading

  1. synapse synapse Public

    Robot VLM and VLA (Vision-Language-Action) inference API helping you manage multimodal prompts, RAG, and location metadata

    Rust 10 1

  2. axon axon Public

    ML platform for training, versioning, and experimenting with VLM and VLA models at scale

    Python 9

  3. dendrite dendrite Public

    ROS2 robot instruction (XRIF) interpreter and task execution system

    C++ 6

  4. simulation-grid simulation-grid Public

    2D simulation environment for reinforcement learning and VLA proof-of-concepts using OpenAI Gym

    Python 5

  5. web web Public

    Web UI for interacting with OpenXRIF services, managing fleets of socially intelligent robots, and running simulations with VLA prompts

    TypeScript 4

  6. cookbook cookbook Public

    Recipes for various robot platforms, action spaces, models, and environments

    Jupyter Notebook 5

Repositories

Showing 8 of 8 repositories

Top languages

Loading…

Most used topics

Loading…