1 code implementation • 25 May 2023 • Guanzhi Wang, Yuqi Xie, Yunfan Jiang, Ajay Mandlekar, Chaowei Xiao, Yuke Zhu, Linxi Fan, Anima Anandkumar
We introduce Voyager, the first LLM-powered embodied lifelong learning agent in Minecraft that continuously explores the world, acquires diverse skills, and makes novel discoveries without human intervention.
1 code implementation • 17 Jun 2022 • Linxi Fan, Guanzhi Wang, Yunfan Jiang, Ajay Mandlekar, Yuncong Yang, Haoyi Zhu, Andrew Tang, De-An Huang, Yuke Zhu, Anima Anandkumar
Autonomous agents have made great strides in specialist domains like Atari games and Go.
6 code implementations • 25 Sep 2020 • Yuke Zhu, Josiah Wong, Ajay Mandlekar, Roberto Martín-Martín, Abhishek Joshi, Soroush Nasiriany, Yifeng Zhu
robosuite is a simulation framework for robot learning powered by the MuJoCo physics engine.
1 code implementation • 10 Jan 2023 • Mayank Mittal, Calvin Yu, Qinxi Yu, Jingzhou Liu, Nikita Rudin, David Hoeller, Jia Lin Yuan, Ritvik Singh, Yunrong Guo, Hammad Mazhar, Ajay Mandlekar, Buck Babich, Gavriel State, Marco Hutter, Animesh Garg
We present Orbit, a unified and modular framework for robot learning powered by NVIDIA Isaac Sim.
1 code implementation • 6 Aug 2021 • Ajay Mandlekar, Danfei Xu, Josiah Wong, Soroush Nasiriany, Chen Wang, Rohun Kulkarni, Li Fei-Fei, Silvio Savarese, Yuke Zhu, Roberto Martín-Martín
Based on the study, we derive a series of lessons including the sensitivity to different algorithmic design choices, the dependence on the quality of the demonstrations, and the variability based on the stopping criteria due to the different objectives in training and evaluation.
1 code implementation • 9 Sep 2019 • Andrey Kurenkov, Ajay Mandlekar, Roberto Martin-Martin, Silvio Savarese, Animesh Garg
The exploration mechanism used by a Deep Reinforcement Learning (RL) agent plays a key role in determining its sample efficiency.
1 code implementation • 20 Oct 2022 • Silviu Pitis, Elliot Creager, Ajay Mandlekar, Animesh Garg
To this end, we show that (1) known local structure in the environment transitions is sufficient for an exponential reduction in the sample complexity of training a dynamics model, and (2) a locally factored dynamics model provably generalizes out-of-distribution to unseen states and actions.
no code implementations • 7 Nov 2018 • Ajay Mandlekar, Yuke Zhu, Animesh Garg, Jonathan Booher, Max Spero, Albert Tung, Julian Gao, John Emmons, Anchit Gupta, Emre Orbay, Silvio Savarese, Li Fei-Fei
Imitation Learning has empowered recent advances in learning robotic manipulation tasks by addressing shortcomings of Reinforcement Learning such as exploration and reward specification.
no code implementations • 11 Nov 2019 • Ajay Mandlekar, Jonathan Booher, Max Spero, Albert Tung, Anchit Gupta, Yuke Zhu, Animesh Garg, Silvio Savarese, Li Fei-Fei
We evaluate the quality of our platform, the diversity of demonstrations in our dataset, and the utility of our dataset via quantitative and qualitative analysis.
no code implementations • 13 Nov 2019 • Ajay Mandlekar, Fabio Ramos, Byron Boots, Silvio Savarese, Li Fei-Fei, Animesh Garg, Dieter Fox
For simple short-horizon manipulation tasks with modest variation in task instances, offline learning from a small set of demonstrations can produce controllers that successfully solve the task.
no code implementations • 13 Mar 2020 • Ajay Mandlekar, Danfei Xu, Roberto Martín-Martín, Silvio Savarese, Li Fei-Fei
In the second stage of GTI, we collect a small set of rollouts from the unconditioned stochastic policy of the first stage, and train a goal-directed agent to generalize to novel start and goal configurations.
no code implementations • 20 Sep 2019 • Dylan P. Losey, Krishnan Srinivasan, Ajay Mandlekar, Animesh Garg, Dorsa Sadigh
Our insight is that we can make assistive robots easier for humans to control by leveraging latent actions.
Robotics
no code implementations • 12 Dec 2020 • Ajay Mandlekar, Danfei Xu, Roberto Martín-Martín, Yuke Zhu, Li Fei-Fei, Silvio Savarese
We develop a simple and effective algorithm to train the policy iteratively on new data collected by the system that encourages the policy to learn how to traverse bottlenecks through the interventions.
no code implementations • 12 Dec 2020 • Albert Tung, Josiah Wong, Ajay Mandlekar, Roberto Martín-Martín, Yuke Zhu, Li Fei-Fei, Silvio Savarese
To address these challenges, we present Multi-Arm RoboTurk (MART), a multi-user data collection platform that allows multiple remote users to simultaneously teleoperate a set of robotic arms and collect demonstrations for multi-arm tasks.
no code implementations • 28 Feb 2021 • Chen Wang, Rui Wang, Ajay Mandlekar, Li Fei-Fei, Silvio Savarese, Danfei Xu
Key to such capability is hand-eye coordination, a cognitive ability that enables humans to adaptively direct their movements at task-relevant objects and be invariant to the objects' absolute spatial location.
no code implementations • 10 Mar 2021 • Samarth Sinha, Ajay Mandlekar, Animesh Garg
Offline reinforcement learning proposes to learn policies from large collected datasets without interacting with the physical environment.
no code implementations • 9 Dec 2021 • Josiah Wong, Albert Tung, Andrey Kurenkov, Ajay Mandlekar, Li Fei-Fei, Silvio Savarese, Roberto Martín-Martín
Doing this is challenging for two reasons: on the data side, current interfaces make collecting high-quality human demonstrations difficult, and on the learning side, policies trained on limited data can suffer from covariate shift when deployed.
no code implementations • 20 Oct 2022 • Soroush Nasiriany, Tian Gao, Ajay Mandlekar, Yuke Zhu
Imitation learning offers a promising path for robots to learn general-purpose behaviors, but traditionally has exhibited limited scalability due to high data supervision requirements and brittle generalization.
no code implementations • 11 Nov 2022 • Kuan Fang, Toki Migimatsu, Ajay Mandlekar, Li Fei-Fei, Jeannette Bohg
ATR selects suitable tasks, which consist of an initial environment state and manipulation goal, for learning robust skills by balancing the diversity and feasibility of the tasks.
no code implementations • 25 May 2023 • Murtaza Dalal, Ajay Mandlekar, Caelan Garrett, Ankur Handa, Ruslan Salakhutdinov, Dieter Fox
In this work, we show that the combination of large-scale datasets generated by TAMP supervisors and flexible Transformer models to fit them is a powerful paradigm for robot manipulation.
no code implementations • 24 Oct 2023 • Ajay Mandlekar, Caelan Garrett, Danfei Xu, Dieter Fox
Finally, we collected 2. 1K demos with HITL-TAMP across 12 contact-rich, long-horizon tasks and show that the system often produces near-perfect agents.
no code implementations • 26 Oct 2023 • Ajay Mandlekar, Soroush Nasiriany, Bowen Wen, Iretiayo Akinola, Yashraj Narang, Linxi Fan, Yuke Zhu, Dieter Fox
Imitation learning from a large set of human demonstrations has proved to be an effective paradigm for building capable robot agents.
no code implementations • 2 Nov 2023 • Shuo Cheng, Caelan Garrett, Ajay Mandlekar, Danfei Xu
Developing intelligent robots for complex manipulation tasks in household and factory settings remains challenging due to long-horizon tasks, contact-rich manipulation, and the need to generalize across a wide variety of object shapes and scene layouts.
no code implementations • 9 Dec 2023 • Motoya Ohnishi, Iretiayo Akinola, Jie Xu, Ajay Mandlekar, Fabio Ramos
As a specific case of our framework, we devise a model predictive control method for path tracking.