This study analyzes semantic segmentation performance across heterogeneously labeled point-cloud datasets relevant to public safety applications, including pre-incident planning systems derived from lidar scans. Using NIST's Point Cloud City dataset (Enfield and Memphis collections), we investigate challenges in unifying differently labeled 3D data. Our methodology employs a graded schema with the KPConv architecture, evaluating performance through IoU metrics on safety-relevant features. Results indicate performance variability: geometrically large objects (e.g. stairs, windows) achieve higher segmentation performance, suggesting potential for navigational context, while smaller safety-critical features exhibit lower recognition rates. Performance is impacted by class imbalance and the limited geometric distinction of smaller objects in typical lidar scans, indicating limitations in detecting certain safety-relevant features using current point-cloud methods. Key identified challenges include insufficient labeled data, difficulties in unifying class labels across datasets, and the need for standardization. Potential directions include automated labeling and multi-dataset learning strategies. We conclude that reliable point-cloud semantic segmentation for public safety necessitates standardized annotation protocols and improved labeling techniques to address data heterogeneity and the detection of small, safety-critical elements.
Audio-driven video generation aims to synthesize realistic videos that align with input audio recordings, akin to the human ability to visualize scenes from auditory input. However, existing approaches predominantly focus on exploring semantic information, such as the classes of sounding sources present in the audio, limiting their ability to generate videos with accurate content and spatial composition. In contrast, we humans can not only naturally identify the semantic categories of sounding sources but also determine their deeply encoded spatial attributes, including locations and movement directions. This useful information can be elucidated by considering specific spatial indicators derived from the inherent physical properties of sound, such as loudness or frequency. As prior methods largely ignore this factor, we present SpA2V, the first framework explicitly exploits these spatial auditory cues from audios to generate videos with high semantic and spatial correspondence. SpA2V decomposes the generation process into two stages: 1) Audio-guided Video Planning: We meticulously adapt a state-of-the-art MLLM for a novel task of harnessing spatial and semantic cues from input audio to construct Video Scene Layouts (VSLs). This serves as an intermediate representation to bridge the gap between the audio and video modalities. 2) Layout-grounded Video Generation: We develop an efficient and effective approach to seamlessly integrate VSLs as conditional guidance into pre-trained diffusion models, enabling VSL-grounded video generation in a training-free manner. Extensive experiments demonstrate that SpA2V excels in generating realistic videos with semantic and spatial alignment to the input audios.
Generative Adversarial Networks (GANs) have achieved realistic super-resolution (SR) of images however, they lack semantic consistency and per-pixel confidence, limiting their credibility in critical remote sensing applications such as disaster response, urban planning and agriculture. This paper introduces Semantic and Uncertainty-Aware ESRGAN (SU-ESRGAN), the first SR framework designed for satellite imagery to integrate the ESRGAN, segmentation loss via DeepLabv3 for class detail preservation and Monte Carlo dropout to produce pixel-wise uncertainty maps. The SU-ESRGAN produces results (PSNR, SSIM, LPIPS) comparable to the Baseline ESRGAN on aerial imagery. This novel model is valuable in satellite systems or UAVs that use wide field-of-view (FoV) cameras, trading off spatial resolution for coverage. The modular design allows integration in UAV data pipelines for on-board or post-processing SR to enhance imagery resulting due to motion blur, compression and sensor limitations. Further, the model is fine-tuned to evaluate its performance on cross domain applications. The tests are conducted on two drone based datasets which differ in altitude and imaging perspective. Performance evaluation of the fine-tuned models show a stronger adaptation to the Aerial Maritime Drone Dataset, whose imaging characteristics align with the training data, highlighting the importance of domain-aware training in SR-applications.
We present a novel approach to multi-agent cooperation by implementing theory of mind (ToM) within active inference. ToM - the ability to understand that others can have differing knowledge and goals - enables agents to reason about others' beliefs while planning their own actions. Unlike previous active inference approaches to multi-agent cooperation, our method neither relies on task-specific shared generative models nor requires explicit communication, while being generalisable. In our framework, the ToM-equipped agent maintains distinct representations of its own and others' beliefs and goals. We extend the sophisticated inference tree-based planning algorithm to systematically explore joint policy spaces through recursive reasoning. Our approach is evaluated through collision avoidance and foraging task simulations. Results demonstrate that ToM-equipped agents cooperate better compared to non-ToM counterparts by being able to avoid collisions and reduce redundant efforts. Crucially, ToM agents accomplish this by inferring others' beliefs solely from observable behaviour. This work advances practical applications in artificial intelligence while providing computational insights into ToM.
Unmanned Aerial Vehicle (UAV) Vision-Language Navigation (VLN) aims to enable agents to accurately localize targets and plan flight paths in complex environments based on natural language instructions, with broad applications in intelligent inspection, disaster rescue, and urban monitoring. Recent progress in Vision-Language Models (VLMs) has provided strong semantic understanding for this task, while reinforcement learning (RL) has emerged as a promising post-training strategy to further improve generalization. However, existing RL methods often suffer from inefficient use of training data, slow convergence, and insufficient consideration of the difficulty variation among training samples, which limits further performance improvement. To address these challenges, we propose \textbf{Semantic-Aware Gaussian Curriculum Scheduling (SA-GCS)}, a novel training framework that systematically integrates Curriculum Learning (CL) into RL. SA-GCS employs a Semantic-Aware Difficulty Estimator (SA-DE) to quantify the complexity of training samples and a Gaussian Curriculum Scheduler (GCS) to dynamically adjust the sampling distribution, enabling a smooth progression from easy to challenging tasks. This design significantly improves training efficiency, accelerates convergence, and enhances overall model performance. Extensive experiments on the CityNav benchmark demonstrate that SA-GCS consistently outperforms strong baselines across all metrics, achieves faster and more stable convergence, and generalizes well across models of different scales, highlighting its robustness and scalability. The implementation of our approach is publicly available.
Multi-object tracking (MOT) enables autonomous vehicles to continuously perceive dynamic objects, supplying essential temporal cues for prediction, behavior understanding, and safe planning. However, conventional tracking-by-detection methods typically rely on static coordinate transformations based on ego-vehicle poses, disregarding ego-vehicle speed-induced variations in observation noise and reference frame changes, which degrades tracking stability and accuracy in dynamic, high-speed scenarios. In this paper, we investigate the critical role of ego-vehicle speed in MOT and propose a Speed-Guided Learnable Kalman Filter (SG-LKF) that dynamically adapts uncertainty modeling to ego-vehicle speed, significantly improving stability and accuracy in highly dynamic scenarios. Central to SG-LKF is MotionScaleNet (MSNet), a decoupled token-mixing and channel-mixing MLP that adaptively predicts key parameters of SG-LKF. To enhance inter-frame association and trajectory continuity, we introduce a self-supervised trajectory consistency loss jointly optimized with semantic and positional constraints. Extensive experiments show that SG-LKF ranks first among all vision-based methods on KITTI 2D MOT with 79.59% HOTA, delivers strong results on KITTI 3D MOT with 82.03% HOTA, and outperforms SimpleTrack by 2.2% AMOTA on nuScenes 3D MOT.
Understanding how language models develop their internal computational structure is a central problem in the science of deep learning. While susceptibilities, drawn from statistical physics, offer a promising analytical tool, their full potential for visualizing network organization remains untapped. In this work, we introduce an embryological approach, applying UMAP to the susceptibility matrix to visualize the model's structural development over training. Our visualizations reveal the emergence of a clear ``body plan,'' charting the formation of known features like the induction circuit and discovering previously unknown structures, such as a ``spacing fin'' dedicated to counting space tokens. This work demonstrates that susceptibility analysis can move beyond validation to uncover novel mechanisms, providing a powerful, holistic lens for studying the developmental principles of complex neural networks.
Paratransit services are vital for individuals who cannot use fixed-route public transit, including those with disabilities. Optimizing these services is essential for transit agencies to deliver high-quality service efficiently. This paper introduces a constraint programming model to jointly optimize route planning and shift scheduling for paratransit operations, along with practical guidance for real-world implementation. A case study in Savannah, Georgia, demonstrates that the new approach is competitive with the state of the art and significantly increases the number of requests served compared to current practices. It is also significantly easier to implement and provides an inherently practical solution for transportation planners. An additional advantage is that the model allows for optimizing shifts without restricting start times to the top of the hour, yielding a further 5% improvement in requests served when applied.
Efficient data collection is essential in applied studies where frequent measurements are costly, time-consuming, or burdensome. This challenge is especially pronounced in functional data settings, where each subject is observed at only a few time points due to practical constraints. Most existing design approaches focus on selecting optimal time points for individual subjects, typically relying on model parameters estimated from a pilot study. However, the design of the pilot study itself has received limited attention. We propose a framework for constructing pilot-study designs that support both accurate trajectory recovery and effective planning of future designs. A search algorithm is developed to generate such high-quality pilot-study designs. Simulation studies and a real data application demonstrate that our approach outperforms commonly used alternatives, highlighting its value in resource-limited settings.
This paper proposes a data-driven motion-planning framework for nonlinear systems that constructs a sequence of overlapping invariant polytopes. Around each randomly sampled waypoint, the algorithm identifies a convex admissible region and solves data-driven linear-matrix-inequality problems to learn several ellipsoidal invariant sets together with their local state-feedback gains. The convex hull of these ellipsoids, still invariant under a piece-wise-affine controller obtained by interpolating the gains, is then approximated by a polytope. Safe transitions between nodes are ensured by verifying the intersection of consecutive convex-hull polytopes and introducing an intermediate node for a smooth transition. Control gains are interpolated in real time via simplex-based interpolation, keeping the state inside the invariant polytopes throughout the motion. Unlike traditional approaches that rely on system dynamics models, our method requires only data to compute safe regions and design state-feedback controllers. The approach is validated through simulations, demonstrating the effectiveness of the proposed method in achieving safe, dynamically feasible paths for complex nonlinear systems.
Accurate link-level bicycling volume estimation is essential for sustainable urban transportation planning. However, many cities face significant challenges of high data sparsity due to limited bicycling count sensor coverage. To address this issue, we propose INSPIRE-GNN, a novel Reinforcement Learning (RL)-boosted hybrid Graph Neural Network (GNN) framework designed to optimize sensor placement and improve link-level bicycling volume estimation in data-sparse environments. INSPIRE-GNN integrates Graph Convolutional Networks (GCN) and Graph Attention Networks (GAT) with a Deep Q-Network (DQN)-based RL agent, enabling a data-driven strategic selection of sensor locations to maximize estimation performance. Applied to Melbourne's bicycling network, comprising 15,933 road segments with sensor coverage on only 141 road segments (99% sparsity) - INSPIRE-GNN demonstrates significant improvements in volume estimation by strategically selecting additional sensor locations in deployments of 50, 100, 200 and 500 sensors. Our framework outperforms traditional heuristic methods for sensor placement such as betweenness centrality, closeness centrality, observed bicycling activity and random placement, across key metrics such as Mean Squared Error (MSE), Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). Furthermore, our experiments benchmark INSPIRE-GNN against standard machine learning and deep learning models in the bicycle volume estimation performance, underscoring its effectiveness. Our proposed framework provides transport planners actionable insights to effectively expand sensor networks, optimize sensor placement and maximize volume estimation accuracy and reliability of bicycling data for informed transportation planning decisions.
Achieving robust cognitive autonomy in robots navigating complex, unpredictable environments remains a fundamental challenge in robotics. This paper presents Underwater Robot Self-Organizing Autonomy (UROSA), a groundbreaking architecture leveraging distributed Large Language Model AI agents integrated within the Robot Operating System 2 (ROS 2) framework to enable advanced cognitive capabilities in Autonomous Underwater Vehicles. UROSA decentralises cognition into specialised AI agents responsible for multimodal perception, adaptive reasoning, dynamic mission planning, and real-time decision-making. Central innovations include flexible agents dynamically adapting their roles, retrieval-augmented generation utilising vector databases for efficient knowledge management, reinforcement learning-driven behavioural optimisation, and autonomous on-the-fly ROS 2 node generation for runtime functional extensibility. Extensive empirical validation demonstrates UROSA's promising adaptability and reliability through realistic underwater missions in simulation and real-world deployments, showing significant advantages over traditional rule-based architectures in handling unforeseen scenarios, environmental uncertainties, and novel mission objectives. This work not only advances underwater autonomy but also establishes a scalable, safe, and versatile cognitive robotics framework capable of generalising to a diverse array of real-world applications.
Clinical trials are a systematic endeavor to assess the safety and efficacy of new drugs or treatments. Conducting such trials typically demands significant financial investment and meticulous planning, highlighting the need for accurate predictions of trial outcomes. Accurately predicting patient enrollment, a key factor in trial success, is one of the primary challenges during the planning phase. In this work, we propose a novel deep learning-based method to address this critical challenge. Our method, implemented as a neural network model, leverages pre-trained language models (PLMs) to capture the complexities and nuances of clinical documents, transforming them into expressive representations. These representations are then combined with encoded tabular features via an attention mechanism. To account for uncertainties in enrollment prediction, we enhance the model with a probabilistic layer based on the Gamma distribution, which enables range estimation. We apply the proposed model to predict clinical trial duration, assuming site-level enrollment follows a Poisson-Gamma process. We carry out extensive experiments on real-world clinical trial data, and show that the proposed method can effectively predict the number of patients enrolled at a number of sites for a given clinical trial, outperforming established baseline models.
Decentralized Multi-Agent Reinforcement Learning (MARL) methods allow for learning scalable multi-agent policies, but suffer from partial observability and induced non-stationarity. These challenges can be addressed by introducing mechanisms that facilitate coordination and high-level planning. Specifically, coordination and temporal abstraction can be achieved through communication (e.g., message passing) and Hierarchical Reinforcement Learning (HRL) approaches to decision-making. However, optimization issues limit the applicability of hierarchical policies to multi-agent systems. As such, the combination of these approaches has not been fully explored. To fill this void, we propose a novel and effective methodology for learning multi-agent hierarchies of message-passing policies. We adopt the feudal HRL framework and rely on a hierarchical graph structure for planning and coordination among agents. Agents at lower levels in the hierarchy receive goals from the upper levels and exchange messages with neighboring agents at the same level. To learn hierarchical multi-agent policies, we design a novel reward-assignment method based on training the lower-level policies to maximize the advantage function associated with the upper levels. Results on relevant benchmarks show that our method performs favorably compared to the state of the art.
Estimating electricity consumption accurately is essential for the planning and operation of energy systems, as well as for billing processes. Standard Load Profiles (SLP) are widely used to estimate consumption patterns of different user groups. However, in Germany these SLP were formulated using historical data from over 20 years ago and have not been adjusted since. Changing electricity consumption behaviour, which leads to increasing deviations between load patterns and SLP, results in a need for a revision taking into account new data. The growing number of smart meters provides a large measurement database, which enables more accurate load modelling. This paper creates updated SLP using recent data. In addition, the assumptions of the SLP method are validated and improvements are proposed, taking into account the ease of applicability. Furthermore, a Fourier Series-based model is proposed as an alternative SLP model. The different models are compared and evaluated.
Distribution grid operation faces new challenges caused by a rising share of renewable energy sources and the introduction of additional types of loads to the grid. With the increasing adoption of distributed generation and emerging prosumer households, Energy Management Systems, which manage and apply flexibility of connected devices, are gaining popularity. While potentially beneficial to grid capacity, strategic energy management also adds to the complexity of distribution grid operation and planning processes. Novel approaches of time-series-based planning likewise face increasingly complex simulation scenarios and rising computational cost. Discrete event modelling helps facilitating simulations of such scenarios by restraining computation to the most relevant points in simulation time. We provide an enhancement of a discrete event distribution grid simulation software that offers fast implementation and testing of energy management algorithms, embedded into a feature-rich simulation environment. Physical models are specified using the Discrete Event System Specification. Furthermore, we contribute a communication protocol that makes use of the discrete event paradigm by only computing flexibility potential when necessary.
Multimodal planning capabilities refer to the ability to predict, reason, and design steps for task execution with multimodal context, which is essential for complex reasoning and decision-making across multiple steps. However, current benchmarks face two key challenges: (1) they cannot directly assess multimodal real-world planning capabilities, and (2) they lack constraints or implicit constraints across modalities. To address these issues, we introduce Multimodal Planning with Complex Constraints (MPCC), the first benchmark to systematically evaluate MLLMs' ability to handle multimodal constraints in planning. To address the first challenge, MPCC focuses on three real-world tasks: Flight Planning, Calendar Planning, and Meeting Planning. To solve the second challenge, we introduce complex constraints (e.g. budget, temporal, and spatial) in these tasks, with graded difficulty levels (EASY, MEDIUM, HARD) to separate constraint complexity from search space expansion. Experiments on 13 advanced MLLMs reveal significant challenges: closed-source models achieve only 21.3% feasible plans, while open-source models average below 11%. Additionally, we observe that MLLMs are highly sensitive to constraint complexity and that traditional multimodal prompting strategies fail in multi-constraint scenarios. Our work formalizes multimodal constraints in planning, provides a rigorous evaluation framework, and highlights the need for advancements in constraint-aware reasoning for real-world MLLM applications.
There is a growing demand for autonomous mobile robots capable of navigating unstructured agricultural environments. Tasks such as weed control in meadows require efficient path planning through an unordered set of coordinates while minimizing travel distance and adhering to curvature constraints to prevent soil damage and protect vegetation. This paper presents an integrated navigation framework combining a global path planner based on the Dubins Traveling Salesman Problem (DTSP) with a Nonlinear Model Predictive Control (NMPC) strategy for local path planning and control. The DTSP generates a minimum-length, curvature-constrained path that efficiently visits all targets, while the NMPC leverages this path to compute control signals to accurately reach each waypoint. The system's performance was validated through comparative simulation analysis on real-world field datasets, demonstrating that the coupled DTSP-based planner produced smoother and shorter paths, with a reduction of about 16% in the provided scenario, compared to decoupled methods. Based thereon, the NMPC controller effectively steered the robot to the desired waypoints, while locally optimizing the trajectory and ensuring adherence to constraints. These findings demonstrate the potential of the proposed framework for efficient autonomous navigation in agricultural environments.
We present 115 compact radio point sources in three galaxies, NGC 5474, NGC 4631 and M51, taken in the most extended (A-)configuration of the Karl G. Jansky Very Large Array at 10GHz. Several of these compact radio point sources have diffuse counterparts identified in previous multi-band studies of resolved radio continuum emission. We find compact counterparts to eight star forming regions, four anomalous microwave emission candidates, and one supernova remnant (SN 2011dh). Nine of the compact radio sources match X-ray counterparts, the majority of which are background galaxies. These AGN are all within the D25 (isophotal diameter) of the host galaxy and might act as contaminants for X-ray binary population studies, highlighting the need for high-resolution multi-band imaging. This study showcases the broad number of science cases that require sensitive radio facilities, like the upcoming Square Kilometre Array and the planned next generation Very Large Array.
Lane segment topology reasoning provides comprehensive bird's-eye view (BEV) road scene understanding, which can serve as a key perception module in planning-oriented end-to-end autonomous driving systems. Existing lane topology reasoning methods often fall short in effectively leveraging temporal information to enhance detection and reasoning performance. Recently, stream-based temporal propagation method has demonstrated promising results by incorporating temporal cues at both the query and BEV levels. However, it remains limited by over-reliance on historical queries, vulnerability to pose estimation failures, and insufficient temporal propagation. To overcome these limitations, we propose FASTopoWM, a novel fast-slow lane segment topology reasoning framework augmented with latent world models. To reduce the impact of pose estimation failures, this unified framework enables parallel supervision of both historical and newly initialized queries, facilitating mutual reinforcement between the fast and slow systems. Furthermore, we introduce latent query and BEV world models conditioned on the action latent to propagate the state representations from past observations to the current timestep. This design substantially improves the performance of temporal perception within the slow pipeline. Extensive experiments on the OpenLane-V2 benchmark demonstrate that FASTopoWM outperforms state-of-the-art methods in both lane segment detection (37.4% v.s. 33.6% on mAP) and centerline perception (46.3% v.s. 41.5% on OLS).
A key challenge in deploying automated vehicles (AVs) is ensuring they make appropriate decisions in ethically challenging everyday driving situations. While much attention has been paid to rare, high-stakes dilemmas such as trolley problems, similar tensions also arise in routine scenarios, such as navigating empty intersections, where multiple human considerations, including legality and comfort, often conflict. Current AV planning systems typically rely on rigid rules, which struggle to balance these competing considerations and can lead to behaviour that misaligns with human expectations. This paper proposes a novel reasons-based trajectory evaluation framework that operationalises the tracking condition of Meaningful Human Control (MHC). The framework models the reasons of human agents, such as regulatory compliance, as quantifiable functions and evaluates how well candidate AV trajectories align with these reasons. By assigning adjustable weights to agent priorities and integrating a balance function to discourage the exclusion of any agent, the framework supports interpretable decision evaluation. Through a real-world-inspired overtaking scenario, we show how this approach reveals tensions, for instance between regulatory compliance, efficiency, and comfort. The framework functions as a modular evaluation layer over existing planning algorithms. It offers a transparent tool for assessing ethical alignment in everyday scenarios and provides a practical step toward implementing MHC in real-world AV deployment.
Vision-Language-Action (VLA) models have demonstrated significant potential in complex scene understanding and action reasoning, leading to their increasing adoption in end-to-end autonomous driving systems. However, the long visual tokens of VLA models greatly increase computational costs. Current visual token pruning methods in Vision-Language Models (VLM) rely on either visual token similarity or visual-text attention, but both have shown poor performance in autonomous driving scenarios. Given that human drivers concentrate on relevant foreground areas while driving, we assert that retaining visual tokens containing this foreground information is essential for effective decision-making. Inspired by this, we propose FastDriveVLA, a novel reconstruction-based vision token pruning framework designed specifically for autonomous driving. FastDriveVLA includes a plug-and-play visual token pruner called ReconPruner, which prioritizes foreground information through MAE-style pixel reconstruction. A novel adversarial foreground-background reconstruction strategy is designed to train ReconPruner for the visual encoder of VLA models. Once trained, ReconPruner can be seamlessly applied to different VLA models with the same visual encoder without retraining. To train ReconPruner, we also introduce a large-scale dataset called nuScenes-FG, consisting of 241K image-mask pairs with annotated foreground regions. Our approach achieves state-of-the-art results on the nuScenes closed-loop planning benchmark across different pruning ratios.
Ethical dilemmas are a common challenge in everyday driving, requiring human drivers to balance competing priorities such as safety, efficiency, and rule compliance. However, much of the existing research in automated vehicles (AVs) has focused on high-stakes "trolley problems," which involve extreme and rare situations. Such scenarios, though rich in ethical implications, are rarely applicable in real-world AV decision-making. In practice, when AVs confront everyday ethical dilemmas, they often appear to prioritise strict adherence to traffic rules. By contrast, human drivers may bend the rules in context-specific situations, using judgement informed by practical concerns such as safety and efficiency. According to the concept of meaningful human control, AVs should respond to human reasons, including those of drivers, vulnerable road users, and policymakers. This work introduces a novel human reasons-based supervision framework that detects when AV behaviour misaligns with expected human reasons to trigger trajectory reconsideration. The framework integrates with motion planning and control systems to support real-time adaptation, enabling decisions that better reflect safety, efficiency, and regulatory considerations. Simulation results demonstrate that this approach could help AVs respond more effectively to ethical challenges in dynamic driving environments by prompting replanning when the current trajectory fails to align with human reasons. These findings suggest that our approach offers a path toward more adaptable, human-centered decision-making in AVs.
In recent times, products have become increasingly complex and highly reliable, so failures typically occur after long periods of operation under normal conditions and may arise from multiple causes. This paper employs simple step-stress partial accelerated life testing (SSSPALT) within the competing risks framework to determine the Bayesian reliability acceptance sampling plan (BRASP) under type-II censoring. Elevating the stress during the life test incurs an additional cost that increases the cost of the life test. In this context, an adaptive scenario is also considered in that sampling plan. The adaptive scenario is as follows: the stress is increased after a certain time if the number of failures up to that point is less than a pre-specified number of failures. The Bayes decision function and Bayes risk are derived for the general loss function. An optimal BRASP under that adaptive SSSPALT is obtained for the quadratic loss function by minimizing Bayes risk. An algorithm is provided to determine the optimal proposed BRASP. Further, comparative studies are conducted between the proposed BRASP, the conventional non-accelerated BRASP, and the conventional accelerated BRASP under type-II censoring to evaluate the effectiveness of the proposed approach. Finally, the methodology is illustrated using real data.
Breast MRI provides high-resolution volumetric imaging critical for tumor assessment and treatment planning, yet manual interpretation of 3D scans remains labor-intensive and subjective. While AI-powered tools hold promise for accelerating medical image analysis, adoption of commercial medical AI products remains limited in low- and middle-income countries due to high license costs, proprietary software, and infrastructure demands. In this work, we investigate whether the Segment Anything Model 2 (SAM2) can be adapted for low-cost, minimal-input 3D tumor segmentation in breast MRI. Using a single bounding box annotation on one slice, we propagate segmentation predictions across the 3D volume using three different slice-wise tracking strategies: top-to-bottom, bottom-to-top, and center-outward. We evaluate these strategies across a large cohort of patients and find that center-outward propagation yields the most consistent and accurate segmentations. Despite being a zero-shot model not trained for volumetric medical data, SAM2 achieves strong segmentation performance under minimal supervision. We further analyze how segmentation performance relates to tumor size, location, and shape, identifying key failure modes. Our results suggest that general-purpose foundation models such as SAM2 can support 3D medical image analysis with minimal supervision, offering an accessible and affordable alternative for resource-constrained settings.
Reconfigurable multi-robot cells offer a promising approach to meet fluctuating assembly demands. However, the recurrent planning of their configurations introduces new challenges, particularly in generating optimized, coordinated multi-robot motion sequences that minimize the assembly duration. This work presents a simulation-based method for generating such optimized sequences. The approach separates assembly steps into task-related core operations and connecting traverse operations. While core operations are constrained and predetermined, traverse operations offer substantial optimization potential. Scheduling the core operations is formulated as an optimization problem, requiring feasible traverse operations to be integrated using a decomposition-based motion planning strategy. Several solution techniques are explored, including a sampling heuristic, tree-based search and gradient-free optimization. For motion planning, a decomposition method is proposed that identifies specific areas in the schedule, which can be solved independently with modified centralized path planning algorithms. The proposed method generates efficient and collision-free multi-robot assembly procedures that outperform a baseline relying on decentralized, robot-individual motion planning. Its effectiveness is demonstrated through simulation experiments.
Tidal disruption events (TDEs) involving supermassive black holes (SMBHs) often exhibit radio emission, yet its physical origin remains uncertain, especially in non-jetted cases. In this Letter, we formulate a general dynamical framework for a radio-emitting shell driven by disk winds and expanding through a power-law ambient medium under the influence of SMBH gravity. We derive and classify power-law-in-time solutions to the governing equations in the adiabatic regime. In particular, a universal $t^{2/3}$ scaling emerges naturally when gravitational energy dominates or is comparable to thermal energy, irrespective of the ambient density profile, whereas the classical Sedov-Taylor solution is recovered when gravity is negligible. Our analysis reveals that, in regimes where SMBH gravity governs the shell expansion, the SMBH mass can be inferred from radio observations of the shell. This approach is independent of and complementary to conventional mass estimators, with direct implications for interpreting radio-emitting TDEs and probing SMBH demographics. Our formalism further predicts that 10-100 GHz monitoring with existing and planned facilities can yield SMBH masses within months of disruption, providing a time-domain analogue to reverberation mapping.
Understanding causal relationships across modalities is a core challenge for multimodal models operating in real-world environments. We introduce ISO-Bench, a benchmark for evaluating whether models can infer causal dependencies between visual observations and procedural text. Each example presents an image of a task step and a text snippet from a plan, with the goal of deciding whether the visual step occurs before or after the referenced text step. Evaluation results on ten frontier vision-language models show underwhelming performance: the best zero-shot F1 is only 0.57, and chain-of-thought reasoning yields only modest gains (up to 0.62 F1), largely behind humans (0.98 F1). Our analysis further highlights concrete directions for improving causal understanding in multimodal models.
The Extract, Transform, Load (ETL) workflow is fundamental for populating and maintaining data warehouses and other data stores accessed by analysts for downstream tasks. A major shortcoming of modern ETL solutions is the extensive need for a human-in-the-loop, required to design and implement context-specific, and often non-generalisable transformations. While related work in the field of ETL automation shows promising progress, there is a lack of solutions capable of automatically designing and applying these transformations. We present FlowETL, a novel example-based autonomous ETL pipeline architecture designed to automatically standardise and prepare input datasets according to a concise, user-defined target dataset. FlowETL is an ecosystem of components which interact together to achieve the desired outcome. A Planning Engine uses a paired input-output datasets sample to construct a transformation plan, which is then applied by an ETL worker to the source dataset. Monitoring and logging provide observability throughout the entire pipeline. The results show promising generalisation capabilities across 14 datasets of various domains, file structures, and file sizes.
We present SMART-Editor, a framework for compositional layout and content editing across structured (posters, websites) and unstructured (natural images) domains. Unlike prior models that perform local edits, SMART-Editor preserves global coherence through two strategies: Reward-Refine, an inference-time rewardguided refinement method, and RewardDPO, a training-time preference optimization approach using reward-aligned layout pairs. To evaluate model performance, we introduce SMARTEdit-Bench, a benchmark covering multi-domain, cascading edit scenarios. SMART-Editor outperforms strong baselines like InstructPix2Pix and HIVE, with RewardDPO achieving up to 15% gains in structured settings and Reward-Refine showing advantages on natural images. Automatic and human evaluations confirm the value of reward-guided planning in producing semantically consistent and visually aligned edits.