An official website of the United States government
continuous state and action spaces is bilevel planning, wherein a high- level search over an abstraction of an environment is used to guide low-level decision-making. Recent work has shown how to enable such bilevel planning by learning abstract models in the form of symbolic operators and neural sam- plers. In this work, we show that existing symbolic operator learning approaches fall short in many robotics domains where a robot’s actions tend to cause a large number of irrelevant changes in the abstract state. This is primarily because they attempt to learn operators that exactly predict all observed changes in the abstract state. To overcome this issue, we propose to learn operators that ‘choose what to predict’ by only modelling changes necessary for abstract planning to achieve specified goals. Experimentally, we show that our approach learns operators that lead to efficient planning across 10 different hybrid robotics domains, including 4 from the challenging BEHAVIOR-100 benchmark, while generalizing to novel initial states, goals, and objects.
more »
« less
Learning where to trust unreliable models in an unstructured world for deformable object manipulation
The world outside our laboratories seldom conforms to the assumptions of our models. This is especially true for dynamics models used in control and motion planning for complex high–degree of freedom systems like deformable objects. We must develop better models, but we must also consider that, no matter how powerful our simulators or how big our datasets, our models will sometimes be wrong. What is more, estimating how wrong models are can be difficult, because methods that predict uncertainty distributions based on training data do not account for unseen scenarios. To deploy robots in unstructured environments, we must address two key questions: When should we trust a model and what do we do if the robot is in a state where the model is unreliable. We tackle these questions in the context of planning for manipulating rope-like objects in clutter. Here, we report an approach that learns a model in an unconstrained setting and then learns a classifier to predict where that model is valid, given a limited dataset of rope-constraint interactions. We also propose a way to recover from states where our model prediction is unreliable. Our method statistically significantly outperforms learning a dynamics function and trusting it everywhere. We further demonstrate the practicality of our method on real-world mock-ups of several domestic and automotive tasks.
more »
« less
- Award ID(s):
- 1750489
- PAR ID:
- 10229750
- Publisher / Repository:
- American Association for the Advancement of Science (AAAS)
- Date Published:
- Journal Name:
- Science Robotics
- Volume:
- 6
- Issue:
- 54
- ISSN:
- 2470-9476
- Page Range / eLocation ID:
- Article No. eabd8170
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Anticipating how a person will interact with objects in an environment is essential for activity understanding, but existing methods are limited to the 2D space of video frames-capturing physically ungrounded predictions of "what" and ignoring the "where" and "how". We introduce FIction for 4D future interaction prediction from videos. Given an input video of a human activity, the goal is to predict which objects at what 3D locations the person will interact with in the next time period (e.g., cabinet, fridge), and how they will execute that interaction (e.g., poses for bending, reaching, pulling). Our novel model FIction fuses the past video observation of the person's actions and their environment to predict both the "where" and "how" of future interactions. Through comprehensive experiments on a variety of activities and real-world environments in EgoExo4D, we show that our proposed approach outperforms prior autoregressive and (lifted) 2D video models substantially, with more than 30% relative gains.more » « less
-
Our brains are, “prediction machines”, where we are continuously comparing our surroundings with predictions from internal models generated by our brains. This is demonstrated by observing our basic low level sensory systems and how they predict environmental changes as we move through space and time. Indeed, even at higher cognitive levels, we are able to do prediction. We can predict how the laws of physics affect people, places, and things and even predict the end of someone’s sentence. In our work, we sought to create an artificial model that is able to mimic early, low level biological predictive behavior in a computer vision system. Our predictive vision model uses spatiotemporal sequence memories learned from deep sparse coding. This model is implemented using a biologically inspired architecture: one that utilizes sequence memories, lateral inhibition, and top-down feed- back in a generative framework. Our model learns the causes of the data in a completely unsupervised manner, by simply observing and learning about the world. Spatiotemporal features are learned by minimizing a reconstruction error convolved over space and time, and can subsequently be used for recognition, classification, and future video prediction. Our experiments show that we are able to accurately predict what will happen in the future; furthermore, we can use our predictions to detect anomalous, unexpected events in both synthetic and real video sequences.more » « less
-
Language models have shown great promise in common-sense related tasks. However, it remains unseen how they would perform in the context of physically situated human-robot interactions, particularly in disaster-relief scenarios. In this paper, we develop a language model evaluation dataset with more than 800 cloze sentences, written to probe for the function of over 200 objects. The sentences are divided into two tasks: an “easy” task where the language model has to choose between vocabulary with different functions (Task 1), and a “challenge” where it has to choose between vocabulary with the same function, yet only one vocabulary item is appropriate given real world constraints on functionality (Task 2). DistilBERT performs with about 80% accuracy for both tasks. To investigate how annotator variability affected those results, we developed a follow-on experiment where we compared our original results with wrong answers chosen based on embedding vector distances. Those results showed increased precision across documents but a 15% decrease in accuracy. We conclude that language models do have a strong knowledge basis for object reasoning, but will require creative fine-tuning strategies in order to be successfully deployed.more » « less
-
Different models can provide differing levels of fidelity when a robot is planning. Analytical models are often fast to evaluate but only work in limited ranges of conditions. Meanwhile, physics simulators are effective at modeling complex interactions between objects but are typically more computationally expensive. Learning when to switch between the various models can greatly improve the speed of planning and task success reliability. In this work, we learn model deviation estimators (MDEs) to predict the error between real-world states and the states outputted by transition models. MDEs can be used to define a model precondition that describes which transitions are accurately modeled. We then propose a planner that uses the learned model preconditions to switch between various models in order to use models in conditions where they are accurate, prioritizing faster models when possible. We evaluate our method on two real-world tasks: placing a rod into a box and placing a rod into a closed drawer.more » « less
