Rose Hendrix

c@rihdrmnasmg.imxeo.ole.

I am a Senior Research Engineer at the Allen Institute for Artificial Intelligence (AI2) on the Perceptual Reasoning and Interaction (PRIOR) team, where I work on embodied AI in unstructured human environments. Prior to that, I earned my PhD in Mechanical Engineering at the University of Washington, where I was advised by Santosh Devasia and Joseph Garbini.

Google Scholar  /  Twitter

profile photo

Research

I'm interested in how to bring robotics into unstructured human environments. Here is my relevant recent work - please see my Google Scholar for a complete history. * denotes equal contribution.

project image

PoliFormer: On-Policy RL with Transformers Results in Masterful Navigators


Kuo-Hao Zeng, Zichen 'Charles' Zhang, Kiana Ehsani, Rose Hendrix, Jordi Salvador, Alvaro Herrasti, Ross Girshick, Aniruddha Kembhavi, Luca Weihs
arXiv, 2024
arxiv / website /

Policy TransFormer (PoliFormer) is a transformer-based policy trained using RL at scale in simulation. PoliFormer achieves SoTA results across LoCoBot and Stretch RE-1, in both simulation and real-world.

project image

Harmonic Mobile Manipulation


Ruihan Yang, Yejin Kim, Rose Hendrix, Aniruddha Kembhavi, Xiaolong Wang, Kiana Ehsani
IROS - Best Paper, Mobile Manipulation, 2024
arxiv / website /

HarmonicMM is an end-to-end learning approach that combines navigation and manipulation, significantly improving success rates in complex tasks like door opening and table cleaning, with successful real-world transfer of agents trained in simulation.

project image

Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Multimodal Models


Matt Detike, Christopher Clark, ..., Rose Hendrix et al.
arXiv, 2024
arxiv / website /

Molmo and PixMo present open weights and open data to advance multimodal AI models, pushing the boundaries of state-of-the-art performance across various tasks.

project image

FLaRe: Achieving Masterful and Adaptive Robot Policies with Large-Scale Reinforcement Learning Fine-Tuning


Jiaheng Hu, Rose Hendrix, Ali Farhadi, Ani Kembhavi, Roberto Martin-Martin, Peter Stone, Kuo-Hao Zeng, Kiana Ehsani
arXiv, 2024
arxiv / website /

FLaRe utilizes large-scale reinforcement learning fine-tuning to create adaptive and highly capable robot policies, achieving state-of-the-art results in both simulated and real-world environments.

project image

Open X-Embodiment: Robotic Learning Datasets and RT-X Models


..., Rose Hendrix, et al.
ICRA - Best Paper, 2024
arxiv / website /

We assemble a dataset from 22 different robots collected through a collaboration between 21 institutions, demonstrating 527 skills (160266 tasks).

project image

SPOC: Imitating Shortest Paths in Simulation Enables Effective Navigation and Manipulation in the Real World


Rose Hendrix*, Kiana Ehsani*, Tanmay Gupta*, Jordi Salvador*, Luca Weihs*, Kuo-Hao Zeng*, Kunal Pratap Singh, Yejin Kim, Winson Han, Alvaro Herrasti, Ranjay Krishna, Dustin Schwenk, Eli VanderBilt, Aniruddha Kembhavi
CVPR, 2024
arxiv / code / website /

We train a supervised model to imitate shortest path trajectories collected from simulation and show that it generalizes to perform effective navigation and manipulation when deployed on real world agents.

project image

Phone2Proc: Bringing Robust Robots Into Our Chaotic World


Rose Hendrix*, Matt Deitke*, Luca Weihs, Ali Farhadi, Kiana Ehsani, Aniruddha Kembhavi
CVPR, 2023
arxiv / code / website /

From a 10-minute iPhone scan of any environment, we condition procedural scene generation on that scan to generate training environments. Training a robot to perform ObjectNav in these scenes dramatically improves sim-to-real performance from 35% to 71% and results in an agent that is remarkably robust to human movement, lighting variations, added clutter, and rearranged objects.





Original design and source code from Jon Barron, modified by Leonid Keselman, and email scrambler by Jeff Donohue.