Exploring the Stratified Space Structure of an RL Game with the Volume Growth Transform

In this work, we explore the structure of the embedding space of a transformer model trained for playing a particular reinforcement learning (RL) game. Specifically, we investigate how a transformer-based Proximal Policy Optimization (PPO) model embeds visual inputs in a simple environment where an agent must collect "coins" while avoiding dynamic obstacles consisting of "spotlights." By adapting Robinson et al.'s study of the volume growth transform for LLMs to the RL setting, we find that the token embedding space for our visual coin collecting game is also not a manifold, and is better modeled as a stratified space, where local dimension can vary from point to point. We further strengthen Robinson's method by proving that fairly general volume growth curves can be realized by stratified spaces. Finally, we carry out an analysis that suggests that as an RL agent acts, its latent representation alternates between periods of low local dimension, while following a fixed sub-strategy, and bursts of high local dimension, where the agent achieves a sub-goal (e.g., collecting an object) or where the environmental complexity increases (e.g., more obstacles appear). Consequently, our work suggests that the distribution of dimensions in a stratified latent space may provide a new geometric indicator of complexity for RL games.

Epistemic Exploration for Generalizable Planning and Learning in Non-Stationary Settings

This paper introduces a new approach for continual planning and model learning in non-stationary stochastic environments expressed using relational representations. Such capabilities are essential for the deployment of sequential decision-making systems in the uncertain, constantly evolving real world. Working in such practical settings with unknown (and non-stationary) transition systems and changing tasks, the proposed framework models gaps in the agent's current state of knowledge and uses them to conduct focused, investigative explorations. Data collected using these explorations is used for learning generalizable probabilistic models for solving the current task despite continual changes in the environment dynamics. Empirical evaluations on several benchmark domains show that this approach significantly outperforms planning and RL baselines in terms of sample complexity in non-stationary settings. Theoretical results show that the system reverts to exhibit desirable convergence properties when stationarity holds.

Indoor and Outdoor 3D Scene Graph Generation via Language-Enabled Spatial Ontologies

This paper proposes an approach to build 3D scene graphs in arbitrary (indoor and outdoor) environments. Such extension is challenging; the hierarchy of concepts that describe an outdoor environment is more complex than for indoors, and manually defining such hierarchy is time-consuming and does not scale. Furthermore, the lack of training data prevents the straightforward application of learning-based tools used in indoor settings. To address these challenges, we propose two novel extensions. First, we develop methods to build a spatial ontology defining concepts and relations relevant for indoor and outdoor robot operation. In particular, we use a Large Language Model (LLM) to build such an ontology, thus largely reducing the amount of manual effort required. Second, we leverage the spatial ontology for 3D scene graph construction using Logic Tensor Networks (LTN) to add logical rules, or axioms (e.g., "a beach contains sand"), which provide additional supervisory signals at training time thus reducing the need for labelled data, providing better predictions, and even allowing predicting concepts unseen at training time. We test our approach in a variety of datasets, including indoor, rural, and coastal environments, and show that it leads to a significant increase in the quality of the 3D scene graph generation with sparsely annotated data.

A perspective on multi-agent communication for information fusion

Collaborative decision making in multi-agent systems typically requires a predefined communication protocol among agents. Usually, agent-level observations are locally processed and information is exchanged using the predefined protocol, enabling the team to perform more efficiently than each agent operating in isolation. In this work, we consider the situation where agents, with complementary sensing modalities must co-operate to achieve a common goal/task by learning an efficient communication protocol. We frame the problem within an actor-critic scheme, where the agents learn optimal policies in a centralized fashion, while taking action in a distributed manner. We provide an interpretation of the emergent communication between the agents. We observe that the information exchanged is not just an encoding of the raw sensor data but is, rather, a specific set of directive actions that depend on the overall task. Simulation results demonstrate the interpretability of the learnt communication in a variety of tasks.

ECO: Egocentric Cognitive Mapping

We present a new method to localize a camera within a previously unseen environment perceived from an egocentric point of view. Although this is, in general, an ill-posed problem, humans can effortlessly and efficiently determine their relative location and orientation and navigate into a previously unseen environments, e.g., finding a specific item in a new grocery store. To enable such a capability, we design a new egocentric representation, which we call ECO (Egocentric COgnitive map). ECO is biologically inspired, by the cognitive map that allows human navigation, and it encodes the surrounding visual semantics with respect to both distance and orientation. ECO possesses three main properties: (1) reconfigurability: complex semantics and geometry is captured via the synthesis of atomic visual representations (e.g., image patch); (2) robustness: the visual semantics are registered in a geometrically consistent way (e.g., aligning with respect to the gravity vector, frontalizing, and rescaling to canonical depth), thus enabling us to learn meaningful atomic representations; (3) adaptability: a domain adaptation framework is designed to generalize the learned representation without manual calibration. As a proof-of-concept, we use ECO to localize a camera within real-world scenes---various grocery stores---and demonstrate performance improvements when compared to existing semantic localization approaches.

Robust Belief Roadmap: Planning Under Intermittent Sensing

In this paper, we extend the recent body of work on planning under uncertainty to include the fact that sensors may not provide any measurement owing to misdetection. This is caused either by adverse environmental conditions that prevent the sensors from making measurements or by the fundamental limitations of the sensors. Examples include RF-based ranging devices that intermittently do not receive the signal from beacons because of obstacles; the misdetection of features by a camera system in detrimental lighting conditions; a LIDAR sensor that is pointed at a glass-based material such as a window, etc. The main contribution of this paper is twofold. We first show that it is possible to obtain an analytical bound on the performance of a state estimator under sensor misdetection occurring stochastically over time in the environment. We then show how this bound can be used in a sample-based path planning algorithm to produce a path that trades off accuracy and robustness. Computational results demonstrate the benefit of the approach and comparisons are made with the state of the art in path planning under state uncertainty.