Taking over arbitrary tasks like humans do with a mobile service robot in open-world settings requires a holistic scene perception for decision-making and high-level control. This paper presents a human-inspired scene perception model to minimize the gap between human and robotic capabilities. The approach takes over fundamental neuroscience concepts, such as a triplet perception split into recognition, knowledge representation, and knowledge interpretation. A recognition system splits the background and foreground to integrate exchangeable image-based object detectors and SLAM, a multi-layer knowledge base represents scene information in a hierarchical structure and offers interfaces for high-level control, and knowledge interpretation methods deploy spatio-temporal scene analysis and perceptual learning for self-adjustment. A single-setting ablation study is used to evaluate the impact of each component on the overall performance for a fetch-and-carry scenario in two simulated and one real-world environment.

Machines that can replicate human intelligence with type 2 reasoning capabilities should be able to reason at multiple levels of spatio-temporal abstractions and scales using internal world models. Devising formalisms to develop such internal world models, which accurately reflect the causal hierarchies inherent in the dynamics of the real world, is a critical research challenge in the domains of artificial intelligence and machine learning. This thesis identifies several limitations with the prevalent use of state space models (SSMs) as internal world models and propose two new probabilistic formalisms namely Hidden-Parameter SSMs and Multi-Time Scale SSMs to address these drawbacks. The structure of graphical models in both formalisms facilitates scalable exact probabilistic inference using belief propagation, as well as end-to-end learning via backpropagation through time. This approach permits the development of scalable, adaptive hierarchical world models capable of representing nonstationary dynamics across multiple temporal abstractions and scales. Moreover, these probabilistic formalisms integrate the concept of uncertainty in world states, thus improving the system's capacity to emulate the stochastic nature of the real world and quantify the confidence in its predictions. The thesis also discuss how these formalisms are in line with related neuroscience literature on Bayesian brain hypothesis and predicitive processing. Our experiments on various real and simulated robots demonstrate that our formalisms can match and in many cases exceed the performance of contemporary transformer variants in making long-range future predictions. We conclude the thesis by reflecting on the limitations of our current models and suggesting directions for future research.

Neural architecture search (NAS) is a challenging problem. Hierarchical search spaces allow for cheap evaluations of neural network sub modules to serve as surrogate for architecture evaluations. Yet, sometimes the hierarchy is too restrictive or the surrogate fails to generalize. We present FaDE which uses differentiable architecture search to obtain relative performance predictions on finite regions of a hierarchical NAS space. The relative nature of these ranks calls for a memory-less, batch-wise outer search algorithm for which we use an evolutionary algorithm with pseudo-gradient descent. FaDE is especially suited on deep hierarchical, respectively multi-cell search spaces, which it can explore by linear instead of exponential cost and therefore eliminates the need for a proxy search space. Our experiments show that firstly, FaDE-ranks on finite regions of the search space correlate with corresponding architecture performances and secondly, the ranks can empower a pseudo-gradient evolutionary search on the complete neural architecture search space.

Counterfactual reasoning, as a crucial manifestation of human intelligence, refers to making presuppositions based on established facts and extrapolating potential outcomes. Existing multimodal large language models (MLLMs) have exhibited impressive cognitive and reasoning capabilities, which have been examined across a wide range of Visual Question Answering (VQA) benchmarks. Nevertheless, how will existing MLLMs perform when faced with counterfactual questions? To answer this question, we first curate a novel \textbf{C}ounter\textbf{F}actual \textbf{M}ulti\textbf{M}odal reasoning benchmark, abbreviated as \textbf{CFMM}, to systematically assess the counterfactual reasoning capabilities of MLLMs. Our CFMM comprises six challenging tasks, each including hundreds of carefully human-labeled counterfactual questions, to evaluate MLLM's counterfactual reasoning capabilities across diverse aspects. Through experiments, interestingly, we find that existing MLLMs prefer to believe what they see, but ignore the counterfactual presuppositions presented in the question, thereby leading to inaccurate responses. Furthermore, we evaluate a wide range of prevalent MLLMs on our proposed CFMM. The significant gap between their performance on our CFMM and that on several VQA benchmarks indicates that there is still considerable room for improvement in existing MLLMs toward approaching human-level intelligence. On the other hand, through boosting MLLMs performances on our CFMM in the future, potential avenues toward developing MLLMs with advanced intelligence can be explored.

We present COmpetitive Mechanisms for Efficient Transfer (COMET), a modular world model which leverages reusable, independent mechanisms across different environments. COMET is trained on multiple environments with varying dynamics via a two-step process: competition and composition. This enables the model to recognise and learn transferable mechanisms. Specifically, in the competition phase, COMET is trained with a winner-takes-all gradient allocation, encouraging the emergence of independent mechanisms. These are then re-used in the composition phase, where COMET learns to re-compose learnt mechanisms in ways that capture the dynamics of intervened environments. In so doing, COMET explicitly reuses prior knowledge, enabling efficient and interpretable adaptation. We evaluate COMET on environments with image-based observations. In contrast to competitive baselines, we demonstrate that COMET captures recognisable mechanisms without supervision. Moreover, we show that COMET is able to adapt to new environments with varying numbers of objects with improved sample efficiency compared to more conventional finetuning approaches.

Embodied reasoning systems integrate robotic hardware and cognitive processes to perform complex tasks typically in response to a natural language query about a specific physical environment. This usually involves changing the belief about the scene or physically interacting and changing the scene (e.g. 'Sort the objects from lightest to heaviest'). In order to facilitate the development of such systems we introduce a new simulating environment that makes use of MuJoCo physics engine and high-quality renderer Blender to provide realistic visual observations that are also accurate to the physical state of the scene. Together with the simulator we propose a new benchmark composed of 10 classes of multi-step reasoning scenarios that require simultaneous visual and physical measurements. Finally, we develop a new modular Closed Loop Interactive Reasoning (CLIER) approach that takes into account the measurements of non-visual object properties, changes in the scene caused by external disturbances as well as uncertain outcomes of robotic actions. We extensively evaluate our reasoning approach in simulation and in the real world manipulation tasks with a success rate above 76% and 64%, respectively.

The thesis develops a view of design in a concept formation framework and outlines a language to describe both the object of the design and the process of designing. The unknown object at the outset of the design work may be seen as an unknown concept that the designer is to define. Throughout the process, she develops a description of this object by relating it to known concepts. The search stops when the designer is satisfied that the design specification is complete enough to satisfy the requirements from it once built. It is then a collection of propositions that all contribute towards defining the design object - a collection of sentences describing relationships between the object and known concepts. Also, the design process itself may be described by relating known concepts - by organizing known abilities into particular patterns of activation, or mobilization. In view of the demands posed to a language to use in this concept formation process, the framework of a Design Process Language (DPL) is developed. The basis for the language are linguistic categories that act as classes of relations used to combine concepts, containing relations used for describing process and object within the same general system, with some relations being process specific, others being object specific, and with the bulk being used both for process and object description. Another outcome is the distinction of modal relations, or relations describing futurity, possibility, willingness, hypothetical events, and the like. The design process almost always includes aspects such as these, and it is thus necessary for a language facilitating design process description to support such relationships to be constructed. The DPL is argued to be a foundation whereupon to build a language that can be used for enabling computers to be more useful - act more intelligently - in the design process.

Text-to-video models have demonstrated substantial potential in robotic decision-making, enabling the imagination of realistic plans of future actions as well as accurate environment simulation. However, one major issue in such models is generalization -- models are limited to synthesizing videos subject to language instructions similar to those seen at training time. This is heavily limiting in decision-making, where we seek a powerful world model to synthesize plans of unseen combinations of objects and actions in order to solve previously unseen tasks in new environments. To resolve this issue, we introduce RoboDreamer, an innovative approach for learning a compositional world model by factorizing the video generation. We leverage the natural compositionality of language to parse instructions into a set of lower-level primitives, which we condition a set of models on to generate videos. We illustrate how this factorization naturally enables compositional generalization, by allowing us to formulate a new natural language instruction as a combination of previously seen components. We further show how such a factorization enables us to add additional multimodal goals, allowing us to specify a video we wish to generate given both natural language instructions and a goal image. Our approach can successfully synthesize video plans on unseen goals in the RT-X, enables successful robot execution in simulation, and substantially outperforms monolithic baseline approaches to video generation.

This paper considers the problem of conflict-free distribution of agents on a circular periphery encompassing all agents. The two key elements of the proposed policy include the construction of a set of convex layers (nested convex polygons) using the initial positions of the agents, and a novel search space region for each of the agents. The search space for an agent on a convex layer is defined as the region enclosed between the lines passing through the agent's position and normal to its supporting edges. Guaranteeing collision-free paths, a goal assignment policy designates a unique goal position within the search space of an agent. In contrast to the existing literature, this work presents a one-shot, collision-free solution to the circular distribution problem by utilizing only the initial positions of the agents. Illustrative examples demonstrate the effectiveness of the proposed policy.

In a hyper-relational knowledge graph (HKG), each fact is composed of a main triple associated with attribute-value qualifiers, which express additional factual knowledge. The hyper-relational knowledge graph completion (HKGC) task aims at inferring plausible missing links in a HKG. Most existing approaches to HKGC focus on enhancing the communication between qualifier pairs and main triples, while overlooking two important properties that emerge from the monotonicity of the hyper-relational graphs representation regime. Stage Reasoning allows for a two-step reasoning process, facilitating the integration of coarse-grained inference results derived solely from main triples and fine-grained inference results obtained from hyper-relational facts with qualifiers. In the initial stage, coarse-grained results provide an upper bound for correct predictions, which are subsequently refined in the fine-grained step. More generally, Qualifier Monotonicity implies that by attaching more qualifier pairs to a main triple, we may only narrow down the answer set, but never enlarge it. This paper proposes the HyperMono model for hyper-relational knowledge graph completion, which realizes stage reasoning and qualifier monotonicity. To implement qualifier monotonicity HyperMono resorts to cone embeddings. Experiments on three real-world datasets with three different scenario conditions demonstrate the strong performance of HyperMono when compared to the SoTA.