Uncertainty maps highlight unreliable regions in segmentation predictions. However, most uncertainty evaluation metrics treat voxels independently, ignoring spatial context and anatomical structure. As a result, they may assign identical scores to qualitatively distinct patterns (e.g., scattered vs. boundary-aligned uncertainty). W e propose three spatially aware metrics that incorporate structural and boundary information and conduct a thorough validation on medical imaging data from the prostate zonal segmentation challenge within the Medical Segmentation Decathlon. Our results demonstrate improved alignment with clinically important factors and better discrimination between meaningful and spurious uncertainty patterns.

As textual reasoning with large language models (LLMs) has advanced significantly, there has been growing interest in enhancing the multimodal reasoning capabilities of large vision-language models (LVLMs). However, existing methods primarily approach multimodal reasoning in a straightforward, text-centric manner, where both reasoning and answer derivation are conducted purely through text, with the only difference being the presence of multimodal input. As a result, these methods often encounter fundamental limitations in spatial reasoning tasks that demand precise geometric understanding and continuous spatial tracking-capabilities that humans achieve through mental visualization and manipulation. To address the limitations, we propose drawing to reason in space, a novel paradigm that enables LVLMs to reason through elementary drawing operations in the visual space. By equipping models with basic drawing operations, including annotating bounding boxes and drawing auxiliary lines, we empower them to express and analyze spatial relationships through direct visual manipulation, meanwhile avoiding the performance ceiling imposed by specialized perception tools in previous tool-integrated reasoning approaches. To cultivate this capability, we develop a three-stage training framework: cold-start training with synthetic data to establish basic drawing abilities, reflective rejection sampling to enhance self-reflection behaviors, and reinforcement learning to directly optimize for target rewards. Extensive experiments demonstrate that our model, named VILASR, consistently outperforms existing methods across diverse spatial reasoning benchmarks, involving maze navigation, static spatial reasoning, video-based reasoning, and multi-view-based reasoning tasks, with an average improvement of 18.4%.

Breast ultrasound imaging is a valuable tool for early breast cancer detection, but automated tumor segmentation is challenging due to inherent noise, variations in scale of lesions, and fuzzy boundaries. To address these challenges, we propose a novel hybrid attention-based network for lesion segmentation. Our proposed architecture integrates a pre-trained DenseNet121 in the encoder part for robust feature extraction with a multi-branch attention-enhanced decoder tailored for breast ultrasound images. The bottleneck incorporates Global Spatial Attention (GSA), Position Encoding (PE), and Scaled Dot-Product Attention (SDPA) to learn global context, spatial relationships, and relative positional features. The Spatial Feature Enhancement Block (SFEB) is embedded at skip connections to refine and enhance spatial features, enabling the network to focus more effectively on tumor regions. A hybrid loss function combining Binary Cross-Entropy (BCE) and Jaccard Index loss optimizes both pixel-level accuracy and region-level overlap metrics, enhancing robustness to class imbalance and irregular tumor shapes. Experiments on public datasets demonstrate that our method outperforms existing approaches, highlighting its potential to assist radiologists in early and accurate breast cancer diagnosis.

Developing autonomous robots capable of learning and reproducing complex motions from demonstrations remains a fundamental challenge in robotics. On the one hand, movement primitives (MPs) provide a compact and modular representation of continuous trajectories. On the other hand, autonomous systems provide control policies that are time independent. We propose in this paper a simple and flexible approach that gathers the advantages of both representations by transforming MPs into autonomous systems. The key idea is to transform the explicit representation of a trajectory as an implicit shape encoded as a distance field. This conversion from a time-dependent motion to a spatial representation enables the definition of an autonomous dynamical system with modular reactions to perturbation. Asymptotic stability guarantees are provided by using Bernstein basis functions in the MPs, representing trajectories as concatenated quadratic Bézier curves, which provide an analytical method for computing distance fields. This approach bridges conventional MPs with distance fields, ensuring smooth and precise motion encoding, while maintaining a continuous spatial representation. By simply leveraging the analytic gradients of the curve and its distance field, a stable dynamical system can be computed to reproduce the demonstrated trajectories while handling perturbations, without requiring a model of the dynamical system to be estimated. Numerical simulations and real-world robotic experiments validate our method's ability to encode complex motion patterns while ensuring trajectory stability, together with the flexibility of designing the desired reaction to perturbations. An interactive project page demonstrating our approach is available at https://mp-df-ds.github.io/.

This note presents a non-associative algebraic framework for the representation and computation of information items in high - dimensional space. This framework is consistent with the principles of spatial computing and with the empirical findings in cognitive science about memory. Computations are performed through a process of multiplication-like binding and non-associative interference-like bundling. Models that rely on associative bundling typically lose order information, which necessitates the use of auxiliary order structures, such as position markers, to represent sequential information that is important for cognitive tasks. In contrast, the non-associative bundling proposed allows the construction of sparse representations of arbitrarily long sequences that maintain their temporal structure across arbitrary lengths. The non-associative nature of the proposed framework results in the representation of a single sequence by two distinct states. The L-state, generated through left-associative bundling, continuously updates and emphasises a recency effect, while the R-state, formed through right-associative bundling, encodes finite sequences or chunks, capturing a primacy effect. The construction of these states may be associated with activity in the prefrontal cortex in relation to short-term memory and hippocampal encoding in long-term memory, respectively. The accuracy of retrieval is contingent upon a decision-making process that is based on the mutual information between the memory states and the cue. The model is able to replicate the Serial Position Curve, which reflects the empirical recency and primacy effects observed in cognitive experiments. Keywords: Memory states, high-dimensional computing (VSA), nonassociative bundling, spatial computing, mutual information, Serial Position Curve T o appear in Neural Computation, V ol 37, Issue 6, June 2025 1 Introduction In essence, the perception of an object is initialised with the activation of a sensory pole. This sensory activation has a rapid decay and lasts for only a few milliseconds.

Capturing human mobility is essential for modeling how people interact with and move through physical spaces, reflecting social behavior, access to resources, and dynamic spatial patterns. To support scalable and transferable analysis across diverse geographies and contexts, there is a need for a generalizable foundation model for spatiotemporal data. While foundation models have transformed language and vision, they remain limited in handling the unique challenges posed by the spatial, temporal, and semantic complexity of mobility data. This vision paper advocates for a new class of spatial foundation models that integrate geolocation semantics with human mobility across multiple scales. Central to our vision is a shift from modeling discrete points of interest to understanding places: dynamic, context-rich regions shaped by human behavior and mobility that may comprise many places of interest. We identify key gaps in adaptability, scalability, and multi-granular reasoning, and propose research directions focused on modeling places and enabling efficient learning. Our goal is to guide the development of scalable, context-aware models for next-generation geospatial intelligence. These models unlock powerful applications ranging from personalized place discovery and logistics optimization to urban planning, ultimately enabling smarter and more responsive spatial decision-making.

--Foundation models have revolutionized robotics by providing rich semantic representations without task-specific training. While many approaches integrate pretrained vision-language models (VLMs) with specialized navigation architectures, the fundamental question remains: can these pretrained embeddings alone successfully guide navigation without additional fine-tuning or specialized modules? We present a minimalist framework that decouples this question by training a behavior cloning policy directly on frozen vision-language embeddings from demonstrations collected by a privileged expert. Our approach achieves a 74% success rate in navigation to language-specified targets, compared to 100% for the state-aware expert, though requiring 3.2 times more steps on average. This performance gap reveals that pretrained embeddings effectively support basic language grounding but struggle with long-horizon planning and spatial reasoning. By providing this empirical baseline, we highlight both the capabilities and limitations of using foundation models as drop-in representations for embodied tasks, offering critical insights for robotics researchers facing practical design tradeoffs between system complexity and performance in resource-constrained scenarios.

Egocentric video-language understanding demands both high efficiency and accurate spatial-temporal modeling. Existing approaches face three key challenges: 1) Excessive pre-training cost arising from multi-stage pre-training pipelines, 2) Ineffective spatial-temporal encoding due to manually split 3D rotary positional embeddings that hinder feature interactions, and 3) Imprecise learning objectives in soft-label multi-instance retrieval, which neglect negative pair correlations. In this paper, we introduce EVA02-AT, a suite of EVA02-based video-language foundation models tailored to egocentric video understanding tasks. EVA02-AT first efficiently transfers an image-based CLIP model into a unified video encoder via a single-stage pretraining. Second, instead of applying rotary positional embeddings to isolated dimensions, we introduce spatial-temporal rotary positional embeddings along with joint attention, which can effectively encode both spatial and temporal information on the entire hidden dimension. This joint encoding of spatial-temporal features enables the model to learn cross-axis relationships, which are crucial for accurately modeling motion and interaction in videos. Third, focusing on multi-instance video-language retrieval tasks, we introduce the Symmetric Multi-Similarity (SMS) loss and a novel training framework that advances all soft labels for both positive and negative pairs, providing a more precise learning objective. Extensive experiments on Ego4D, EPIC-Kitchens-100, and Charades-Ego under zero-shot and fine-tuning settings demonstrate that EVA02-AT achieves state-of-the-art performance across diverse egocentric video-language tasks with fewer parameters. Models with our SMS loss also show significant performance gains on multi-instance retrieval benchmarks. Our code and models are publicly available at https://github.com/xqwang14/EVA02-AT .

Predicting individuals' next locations is a core task in human mobility modelling, with wide-ranging implications for urban planning, transportation, public policy and personalised mobility services. Traditional approaches largely depend on location embeddings learned from historical mobility patterns, limiting their ability to encode explicit spatial information, integrate rich urban semantic context, and accommodate previously unseen locations. To address these challenges, we explore the application of CaLLiPer -- a multimodal representation learning framework that fuses spatial coordinates and semantic features of points of interest through contrastive learning -- for location embedding in individual mobility prediction. CaLLiPer's embeddings are spatially explicit, semantically enriched, and inductive by design, enabling robust prediction performance even in scenarios involving emerging locations. Through extensive experiments on four public mobility datasets under both conventional and inductive settings, we demonstrate that CaLLiPer consistently outperforms strong baselines, particularly excelling in inductive scenarios. Our findings highlight the potential of multimodal, inductive location embeddings to advance the capabilities of human mobility prediction systems. We also release the code and data (https://github.com/xlwang233/Into-the-Unknown) to foster reproducibility and future research.

Living environments play a vital role in the prevalence and progression of diseases, and understanding their impact on patient's health status becomes increasingly crucial for developing AI models. However, due to the lack of long-term and fine-grained spatial and temporal data in public and population health studies, most existing studies fail to incorporate environmental data, limiting the models' performance and real-world application. To address this shortage, we developed SatHealth, a novel dataset combining multimodal spatiotemporal data, including environmental data, satellite images, all-disease prevalences estimated from medical claims, and social determinants of health (SDoH) indicators. We conducted experiments under two use cases with SatHealth: regional public health modeling and personal disease risk prediction. Experimental results show that living environmental information can significantly improve AI models' performance and temporal-spatial generalizability on various tasks. Finally, we deploy a web-based application to provide an exploration tool for SatHealth and one-click access to both our data and regional environmental embedding to facilitate plug-and-play utilization. SatHealth is now published with data in Ohio, and we will keep updating SatHealth to cover the other parts of the US. With the web application and published code pipeline, our work provides valuable angles and resources to include environmental data in healthcare research and establishes a foundational framework for future research in environmental health informatics.