Vinicius G. Goecks

Army Research Laboratory | ALC · Human Research and Engineering Directorate (HRED)

Traditionally, learning from human demonstrations via direct behavior cloning can lead to high-performance policies given that the algorithm has access to large amounts of high-quality data covering the most likely scenarios to be encountered when the agent is operating. However, in real-world scenarios, expert data is limited and it is desired to...

We held the first-ever MineRL Benchmark for Agents that Solve Almost-Lifelike Tasks (MineRL BASALT) Competition at the Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS 2021). The goal of the competition was to promote research towards agents that use learning from human feedback (LfHF) techniques to solve open-world tasks....

Games and simulators can be a valuable platform to execute complex multi-agent, multiplayer, imperfect information scenarios with significant parallels to military applications: multiple participants manage resources and make decisions that command assets to secure specific areas of a map or neutralize opposing forces. These characteristics have at...

Real-world tasks of interest are generally poorly defined by human-readable descriptions and have no pre-defined reward signals unless it is defined by a human designer. Conversely, data-driven algorithms are often designed to solve a specific, narrowly defined, task with performance metrics that drives the agent's learning. In this work, we presen...

In the field of human-robot interaction, teaching learning agents from human demonstrations via supervised learning has been widely studied and successfully applied to multiple domains such as self-driving cars and robot manipulation. However, the majority of the work on learning from human demonstrations utilizes only behavioral information from t...

Recent successes combine reinforcement learning algorithms and deep neural networks, despite reinforcement learning not being widely applied to robotics and real world scenarios. This can be attributed to the fact that current state-of-the-art, end-to-end reinforcement learning approaches still require thousands or millions of data samples to conve...

This paper investigates how to efficiently transition and update policies, trained initially with demonstrations, using off-policy actor-critic reinforcement learning. It is well-known that techniques based on Learning from Demonstrations, for example behavior cloning, can lead to proficient policies given limited data. However, it is currently unc...

Advances in machine learning and deep neural networks for object detection, coupled with lower cost and power requirements of cameras, led to promising vision-based solutions for sUAS detection. However, solely relying on the visible spectrum has previously led to reliability issues in low contrast scenarios such as sUAS flying below the treeline a...

Learning from demonstration has been widely studied in machine learning but becomes challenging when the demonstrated trajectories are unstructured and follow different objectives. Our proposed work, Plannable Option Discovery Network (PODNet), addresses how to segment an unstructured set of demonstrated trajectories for option discovery. This enab...

This paper investigates how to utilize different forms of human interaction to safely train autonomous systems in realtime by learning from both human demonstrations and interventions. We implement two components of the Cycle-of Learning for Autonomous Systems, which is our framework for combining multiple modalities of human interaction. The curre...

This paper investigates how to utilize different forms of human interaction to safely train autonomous systems in real-time by learning from both human demonstrations and interventions. We implement two components of the Cycle-of-Learning for Autonomous Systems, which is our framework for combining multiple modalities of human interaction. The curr...

We discuss different types of human-robot interaction paradigms in the context of training end-to-end reinforcement learning algorithms. We provide a taxonomy to categorize the types of human interaction and present our Cycle-of-Learning framework for autonomous systems that combines different human-interaction modalities with reinforcement learnin...

Recent developments in artificial intelligence enabled training of autonomous robots without human supervision. Even without human supervision during training, current models have yet to be human-engineered and have neither guarantees to match human expectation nor perform within safety bounds. This paper proposes CyberSteer to leverage human-robot...

Refuelling in cis-lunar space can enable more massive payloads to be launched from Earth to deep space, significantly increasing access for scientific and industrial space applications on at the Moon, Mars, and beyond. The 2017 Caltech Space Challenge saw two teams of graduate and undergraduate students participate in a mission design competition t...

This paper presents a new approach for enhanced 3D situational awareness during astronaut controlled robotics operations in space. The goal is to improve crew efficiency and assure collision avoidance, while simplifying operations. The platform was developed to aid astronauts control the Mobile Servicing System (MSS) Remote Manipulator System (RMS)...

Recent technological developments surrounding CubeSats and Commercial Off-The-Shelf space hardware have drastically reduced the cost of producing and flying a satellite mission. As the barriers to entry fall, space missions become a viable option for more students and research groups. Many of these missions require accurate spacecraft pointing and...

We present a method for performing low cost attitude estimation for CubeSat type missions. Our algorithm uses measurements from a custom built sun sensor, a star camera, and inertial measurements. These sensing measurements are supplied in real-time to an Multiplicative Kalman Filter for the purpose of generating continuous attitude estimates. The...