Kevin Zakka Claim this profile

Isaac Lab API, powered by MuJoCo-Warp, for RL and robotics research.

Python inverse kinematics based on MuJoCo.

[CoRL '23] Dexterous piano playing with deep reinforcement learning.

A collection of high-quality models for the MuJoCo physics engine, curated by DeepMind.

Multi-Joint dynamics with Contact. A general purpose physics simulator.

A Tensorflow implementation of Spatial Transformer Networks.

PyTorch Boilerplate For Research

A PyTorch Implementation of "Recurrent Models of Visual Attention"

Minimal, clean, single-file implementations of common robotics controllers in MuJoCo.

A PyTorch implementation of Neighbourhood Components Analysis.

A CLI for processing composite Wavefront OBJ files for use in MuJoCo.

MuJoCo Models for Google's Scanned Objects Dataset

An ever-growing playground of notebooks showcasing CLIP's impressive zero-shot capabilities

Research boilerplate for PyTorch.

Python Wrapper for t-SNE Visualization

A PyTorch implementation of Implicit Behavioral Cloning

[ICRA 2020] Train generalizable policies for kit assembly with self-supervised dense correspondence learning.

My blog's code repository.

Learning some numerical linear algebra.

A browser-based 3D viewer for MuJoCo

RL code for training piano-playing policies for RoboPianist.

[CoRL 2021] A robotics benchmark for cross-embodiment imitation.

Benchmarking PyTorch variants of TSDF fusion.

A 3x3x3 puzzle cube model for MuJoCo.

No description.

Learning about RANSAC.

Standalone library of frequently-used wrappers for dm_env environments.

A tiny reinforcement learning codebase for continuous control, built on top of JAX.

My open-source software contributions.

A curated list of awesome robot descriptions

No description.

Import robot descriptions as Python modules

Learning about volumetric fusion.

Learning about various image blurring techniques.

Customizable Reinforcement Learning

No description.

No description.

RL training visualizer for Mujoco Playground environments + Brax training

Fast and simple implementation of RL algorithms, designed to run fully on GPU.

GPU-optimized version of the MuJoCo physics simulator, designed for NVIDIA hardware.

A collection of tasks built on mjlab

No description.

Massively parallel rigidbody physics simulation on accelerator hardware.

Toy implementations of algorithms for path planning, system modelling, optimal control, estimation, and more.

Web-based 3D visualization + Python

No description.

No description.

Ubisoft La Forge - Animation Dataset

No description.

Box of handy tools for Sphinx 🧰 📔

C++ and Python SDK for ARX5 robot arm develeoped by @yihuai-gao (@real-stanford)

Google DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.

This code corresponds to simulation environments used as part of the MimicGen project.

No description.

HACMan: Learning Hybrid Actor-Critic Maps for 6D Non-Prehensile Manipulation

Project Page for Lifelike Agility and Play in Quadrupedal Robots using Reinforcement Learning and Generative Pre-trained Models

dataloading is my passion

robomimic: A Modular Framework for Robot Learning from Demonstration

My robotics research toolkit.