Search

The contributions of this short technical note are two-fold. Firstly, we introduce a modified version of a generalized Bellman-Ford algorithm calculating the value function of optimal control problems defined on hyper-graphs. Those Bellman-Ford algorithms can be used in particular for the synthesis of near-optimal controllers by the principle of symbolic control. Our modification causes less nodes of the hyper-graph being iterated during the execution compared to our initial version of the algorithm published in 2020. Our second contribution lies in the field of Plan recognition applied to drone missions driven by symbolic controllers. We address and resolve the Plan and Goal Recognition monitor's dependence on a pre-defined initial guess for a drone's task allocation and mission execution. To validate the enhanced implementation, we use a more challenging scenario for UAV-based aerial firefighting, demonstrating the practical applicability and robustness of the system architecture.

Quantum interference is known to become extinct with distinguishing information, as illustrated by the ubiquitous double-slit experiment or the two-photon HOM effect. In the former case single particle interference is destroyed with which-path information while in the latter bunching interference tails-off as photons become distinguishable. It has been observed that when more than two particles are involved, these interference patterns are in general a non monotonic function of the distinguishability. Here we perform a comprehensive characterization, both theoretically and experimentally, of four-photon interference by analyzing the corresponding correlation functions, contemplating several degrees of distinguishability across different parameters. This study provides all the necessary tools to quantify the impact of multi-photon interference on precision measurements of parameters such as phase, frequency, and time difference. We apply these insights to quantify the precision in the estimation of an interferometric phase in a two-port interferometer using a four-photon state. Our results reveal that, for certain phase values, partially distinguishable multi-photon states can achieve higher Fisher information values compared to the two-photon experiment. These findings highlight the potential of distinguishable multi-photon states for enhanced precision in quantum metrology and related applications., Comment: 12 pages, 6 figures

Deep neural networks (DNNs) remain challenged by distribution shifts in complex open-world domains like automated driving (AD): Absolute robustness against yet unknown novel objects (semantic shift) or styles like lighting conditions (covariate shift) cannot be guaranteed. Hence, reliable operation-time monitors for identification of out-of-training-data-distribution (OOD) scenarios are imperative. Current approaches for OOD classification are untested for complex domains like AD, are limited in the kinds of shifts they detect, or even require supervision with OOD samples. To prepare for unanticipated shifts, we instead establish a framework around a principled, unsupervised, and model-agnostic method that unifies detection of all kinds of shifts: Find a full model of the training data's feature distribution, to then use its density at new points as in-distribution (ID) score. To implement this, we propose to combine the newly available Vision Foundation Models (VFM) as feature extractors with one of four alternative density modeling techniques. In an extensive benchmark of 4 VFMs against 20 baselines, we show the superior performance of VFM feature encodings compared to shift-specific OOD monitors. Additionally, we find that sophisticated architectures outperform larger latent space dimensionality; and our method identifies samples with higher risk of errors on downstream tasks, despite being model-agnostic. This suggests that VFMs are promising to realize model-agnostic, unsupervised, reliable safety monitors in complex vision tasks.

Quantum error correction is needed for quantum computers to be capable of fault-tolerantly executing algorithms using hundreds of logical qubits. Recent experiments have demonstrated subthreshold error rates for state preservation of a single logical qubit. In addition, the realization of universal quantum computation requires the implementation of logical entangling gates. Lattice surgery offers a practical approach for implementing such gates, particularly in planar quantum processor layouts. In this work, we demonstrate lattice surgery between two distance-three repetition-code qubits by splitting a single distance-three surface-code qubit. Using a quantum circuit fault-tolerant to bit-flip errors, we achieve an improvement in the value of the decoded $ZZ$ logical two-qubit observable compared to a similar non-encoded circuit. By preparing the surface-code qubit in initial states parametrized by a varying polar angle, we evaluate the performance of the lattice surgery operation for non-cardinal states on the logical Bloch sphere and employ logical two-qubit tomography to reconstruct the Pauli transfer matrix of the operation. In this way, we demonstrate the functional building blocks needed for lattice surgery operations on larger-distance codes based on superconducting circuits., Comment: 19 pages, 13 figures

Event cameras are neuromorphically inspired sensors that sparsely and asynchronously report brightness changes. Their unique characteristics of high temporal resolution, high dynamic range, and low power consumption make them well-suited for addressing challenges in monocular depth estimation (e.g., high-speed or low-lighting conditions). However, current existing methods primarily treat event streams as black-box learning systems without incorporating prior physical principles, thus becoming over-parameterized and failing to fully exploit the rich temporal information inherent in event camera data. To address this limitation, we incorporate physical motion principles to propose an interpretable monocular depth estimation framework, where the likelihood of various depth hypotheses is explicitly determined by the effect of motion compensation. To achieve this, we propose a Focus Cost Discrimination (FCD) module that measures the clarity of edges as an essential indicator of focus level and integrates spatial surroundings to facilitate cost estimation. Furthermore, we analyze the noise patterns within our framework and improve it with the newly introduced Inter-Hypotheses Cost Aggregation (IHCA) module, where the cost volume is refined through cost trend prediction and multi-scale cost consistency constraints. Extensive experiments on real-world and synthetic datasets demonstrate that our proposed framework outperforms cutting-edge methods by up to 10\% in terms of the absolute relative error metric, revealing superior performance in predicting accuracy., Comment: 9 pages, 3 figures

Geological materials are often seen as the antithesis of soft; rocks are hard. However, during the formation of minerals and rocks, all the systems we shall discuss, indeed geological materials in general, pass through a stage where they are soft. This occurs either because they are at a high temperature -- igneous or metamorphic rock -- or because they are at a lower temperature but in the presence of water -- sedimentary rock. For this reason it is useful to introduce soft-matter concepts into the geological domain. There is a universality in the diverse instances of geological patterns that may be appreciated by looking at the common aspect in their formation of having passed through a stage as soft matter.

Ultrasound localization microscopy (ULM) enables microvascular imaging at spatial resolutions beyond the acoustic diffraction limit, offering significant clinical potentials. However, ULM performance relies heavily on microbubble (MB) signal sparsity, the number of detected MBs, and signal-to-noise ratio (SNR), all of which vary in clinical scenarios involving bolus MB injections. These sources of variations underscore the need to optimize MB dosage, data acquisition timing, and imaging settings in order to standardize and optimize ULM of microvasculature. This pilot study investigated temporal changes in MB signals during bolus injections in both pig and human models to optimize data acquisition for clinical ULM. Quantitative indices were developed to evaluate MB signal quality, guiding selection of acquisition timing that balances the MB localization quality and adequate MB counts. The effects of transmitted voltage and dosage were also explored. In the pig model, a relatively short window (approximately 10 seconds) for optimal acquisition was identified during the rapid wash-out phase, highlighting the need for real-time MB signal monitoring during data acquisition. The slower wash-out phase in humans allowed for a more flexible imaging window of 1-2 minutes, while trade-offs were observed between localization quality and MB density (or acquisition length) at different wash-out phase timings. Guided by these findings, robust ULM imaging was achieved in both pig and human kidneys using a short period of data acquisition, demonstrating its feasibility in clinical practice. This study provides insights into optimizing data acquisition for consistent and reproducible ULM, paving the way for its standardization and broader clinical applications., Comment: 33 pages, 9 figures

Bayesian inference in function space has gained attention due to its robustness against overparameterization in neural networks. However, approximating the infinite-dimensional function space introduces several challenges. In this work, we discuss function space inference via particle optimization and present practical modifications that improve uncertainty estimation and, most importantly, make it applicable for large and pretrained networks. First, we demonstrate that the input samples, where particle predictions are enforced to be diverse, are detrimental to the model performance. While diversity on training data itself can lead to underfitting, the use of label-destroying data augmentation, or unlabeled out-of-distribution data can improve prediction diversity and uncertainty estimates. Furthermore, we take advantage of the function space formulation, which imposes no restrictions on network parameterization other than sufficient flexibility. Instead of using full deep ensembles to represent particles, we propose a single multi-headed network that introduces a minimal increase in parameters and computation. This allows seamless integration to pretrained networks, where this repulsive last-layer ensemble can be used for uncertainty aware fine-tuning at minimal additional cost. We achieve competitive results in disentangling aleatoric and epistemic uncertainty for active learning, detecting out-of-domain data, and providing calibrated uncertainty estimates under distribution shifts with minimal compute and memory.

The emergence of Software-Defined Vehicles (SDVs) signifies a shift from a distributed network of electronic control units (ECUs) to a centralized computing architecture within the vehicle's electrical and electronic systems. This transition addresses the growing complexity and demand for enhanced functionality in traditional E/E architectures, with containerization and virtualization streamlining software development and updates within the SDV framework. While widely used in cloud computing, their performance and suitability for intelligent vehicles have yet to be thoroughly evaluated. In this work, we conduct a comprehensive performance evaluation of containerization and virtualization on embedded and high-performance AMD64 and ARM64 systems, focusing on CPU, memory, network, and disk metrics. In addition, we assess their impact on real-world automotive applications using the Autoware framework and further integrate a microservice-based architecture to evaluate its start-up time and resource consumption. Our extensive experiments reveal a slight 0-5% performance decline in CPU, memory, and network usage for both containerization and virtualization compared to bare-metal setups, with more significant reductions in disk operations-5-15% for containerized environments and up to 35% for virtualized setups. Despite these declines, experiments with actual vehicle applications demonstrate minimal impact on the Autoware framework, and in some cases, a microservice architecture integration improves start-up time by up to 18%., Comment: 15 pages, 15 figures

Cardiovascular magnetic resonance imaging is a powerful diagnostic tool for assessing cardiac structure and function. Traditional breath-held imaging protocols, however, pose challenges for patients with arrhythmias or limited breath-holding capacity. We introduce Motion-Guided Deep Image prior (M-DIP), a novel unsupervised reconstruction framework for accelerated real-time cardiac MRI. M-DIP employs a spatial dictionary to synthesize a time-dependent template image, which is further refined using time-dependent deformation fields that model cardiac and respiratory motion. Unlike prior DIP-based methods, M-DIP simultaneously captures physiological motion and frame-to-frame content variations, making it applicable to a wide range of dynamic applications. We validate M-DIP using simulated MRXCAT cine phantom data as well as free-breathing real-time cine and single-shot late gadolinium enhancement data from clinical patients. Comparative analyses against state-of-the-art supervised and unsupervised approaches demonstrate M-DIP's performance and versatility. M-DIP achieved better image quality metrics on phantom data, as well as higher reader scores for in-vivo patient data.

3D Gaussian Splatting (3DGS) allows flexible adjustments to scene representation, enabling continuous optimization of scene quality during dense visual simultaneous localization and mapping (SLAM) in static environments. However, 3DGS faces challenges in handling environmental disturbances from dynamic objects with irregular movement, leading to degradation in both camera tracking accuracy and map reconstruction quality. To address this challenge, we develop an RGB-D dense SLAM which is called Gaussian Splatting SLAM in Dynamic Environments (Gassidy). This approach calculates Gaussians to generate rendering loss flows for each environmental component based on a designed photometric-geometric loss function. To distinguish and filter environmental disturbances, we iteratively analyze rendering loss flows to detect features characterized by changes in loss values between dynamic objects and static components. This process ensures a clean environment for accurate scene reconstruction. Compared to state-of-the-art SLAM methods, experimental results on open datasets show that Gassidy improves camera tracking precision by up to 97.9% and enhances map quality by up to 6%., Comment: This paper is currently under reviewed for ICRA 2025

In this paper, we explore the integration of Large Language Models (LLMs) with Retrieval-Augmented Generation (RAG) to enhance automated design and software development in the automotive industry. We present two case studies: a standardization compliance chatbot and a design copilot, both utilizing RAG to provide accurate, context-aware responses. We evaluate four LLMs-GPT-4o, LLAMA3, Mistral, and Mixtral -- comparing their answering accuracy and execution time. Our results demonstrate that while GPT-4 offers superior performance, LLAMA3 and Mistral also show promising capabilities for local deployment, addressing data privacy concerns in automotive applications. This study highlights the potential of RAG-augmented LLMs in improving design workflows and compliance in automotive engineering., Comment: Conference paper accepted in IEEE FLLM 2024

This paper presents a monitoring framework that infers the level of autonomous vehicle (AV) collision risk based on its object detector's performance using only monocular camera images. Essentially, the framework takes two sets of predictions produced by different algorithms and associates their inconsistencies with the collision risk via fuzzy inference. The first set of predictions is obtained through retrieving safety-critical 2.5D objects from a depth map, and the second set comes from the AV's 3D object detector. We experimentally validate that, based on Intersection-over-Union (IoU) and a depth discrepancy measure, the inconsistencies between the two sets of predictions strongly correlate to the safety-related error of the 3D object detector against ground truths. This correlation allows us to construct a fuzzy inference system and map the inconsistency measures to an existing collision risk indicator. In particular, we apply various knowledge- and data-driven techniques and find using particle swarm optimization that learns general fuzzy rules gives the best mapping result. Lastly, we validate our monitor's capability to produce relevant risk estimates with the large-scale nuScenes dataset and show it can safeguard an AV in closed-loop simulations., Comment: 7 pages (IEEE double column format), 5 figures, 3 tables, submitted to ICRA 2025

Across many plasma applications, the underlying phenomena and interactions among the involved processes are known to exhibit three-dimensional characteristics. Furthermore, the global properties and evolution of plasma systems are often determined by a process called inverse energy cascade, where kinetic plasma processes at the microscopic scale interact and lead to macroscopic coherent structures. These structures can have a major impact on the stability of plasma discharges, with detrimental effects on the operation and performance of plasma technologies. Kinetic particle-in-cell (PIC) methods offer a sufficient level of fidelity to capture these processes and behaviors. However, three-dimensional PIC simulations that can cost-effectively overcome the curse of dimensionality and enable full-scale simulations of real-world time significance have remained elusive. Tackling the enormous computational cost issue associated with conventional PIC schemes, the computationally efficient reduced-order (RO) PIC approach provides a viable path to 3D simulations of real-size plasma systems. This part II paper builds upon the improvements to the RO-PIC's underpinning formulation discussed in part I and extends the novel "first-order" RO-PIC formulation to 3D. The resulting Quasi-3D (Q3D) implementation is rigorously verified in this paper, both at the module level of the Q3D reduced-dimension Poisson solver (RDPS) and at the global PIC code level. The plasma test cases employed correspond to 3D versions of the 2D configurations studied in Part I, including a 3D extension to the Diocotron instability problem. The detailed verifications of the Q3D RO-PIC confirm that it maintains the expected levels of cost-efficiency and accuracy, demonstrating the ability of the approach to indistinguishably reproduce full-3D simulation results at a fraction of the computational cost., Comment: 24 pages, 21 figures

The particle-in-cell (PIC) method is a well-established and widely used kinetic plasma modelling approach that provides a hybrid Lagrangian-Eulerian approach to solve the plasma kinetic equation. Despite its power in capturing details of the underlying physics of plasmas, conventional PIC implementations are associated with a significant computational cost, rendering their applications for real-world plasma science and engineering challenges impractical. The acceleration of the PIC method has thus become a topic of high interest, with several approaches having been pursued to this end. Among these, the concept of reduced-order (RO) PIC simulations, first introduced in 2023, provides a uniquely flexible and computationally efficient framework for kinetic plasma modelling - characteristics verified extensively in various plasma configurations. In this two-part article, we report the latest progress achieved on RO-PIC. Part I article revisits the original RO-PIC formulation and introduces refinements that substantially enhance the cost-efficiency and accuracy of the method. We discuss these refinements in comparison against the original formulation, illustrating the progression to a "first-order" implementation from the baseline "zeroth-order" one. In a detailed step-by-step verification, we first test the newly updated reduced-dimension Poisson solver (RDPS) in the first-order RO-PIC against its zeroth-order counterpart using test-case Poisson problems. Next, comparing against the zeroth-order version, we examine the performance of the complete first-order RO-PIC code in two-dimensional plasma problems. The detailed verifications demonstrate that the improvements in the RO-PIC formulation enable the approach to provide full-2D-equivalent results at a substantially lower (up to an order of magnitude) computational cost compared to the zeroth-order RO-PIC., Comment: 29 pages, 24 figures

Recent advancements in robotics have transformed industries such as manufacturing, logistics, surgery, and planetary exploration. A key challenge is developing efficient motion planning algorithms that allow robots to navigate complex environments while avoiding collisions and optimizing metrics like path length, sweep area, execution time, and energy consumption. Among the available algorithms, sampling-based methods have gained the most traction in both research and industry due to their ability to handle complex environments, explore free space, and offer probabilistic completeness along with other formal guarantees. Despite their widespread application, significant challenges still remain. To advance future planning algorithms, it is essential to review the current state-of-the-art solutions and their limitations. In this context, this work aims to shed light on these challenges and assess the development and applicability of sampling-based methods. Furthermore, we aim to provide an in-depth analysis of the design and evaluation of ten of the most popular planners across various scenarios. Our findings highlight the strides made in sampling-based methods while underscoring persistent challenges. This work offers an overview of the important ongoing research in robotic motion planning., Comment: 20 pages, 11 figures

We compute the MU-based syntomic cohomologies, mod $(p,v_1,\cdots,v_{n+1})$, of all $\mathbb{E}_1$-MU-algebra forms of connective Morava K-theory k(n). As qualitative consequences, we deduce the Lichtenbaum--Quillen conjecture, telescope conjecture, and redshift conjecture for the algebraic K-theories of all $\mathbb{E}_{1}$-$\mathbb{S}$-algebra forms of $(2p^n-2)$-periodic Morava K-theory. Notably, the motivic spectral sequence computing $\pi_*TC(k(n))_p$ is concentrated on at most three lines, independently of $n$., Comment: 63 pages, 5 figures, Comments welcome!

Deep learning is actively being used in biometrics to develop efficient identification and verification systems. Handwritten signatures are a common subset of biometric data for authentication purposes. Generative adversarial networks (GANs) learn from original and forged signatures to generate forged signatures. While most GAN techniques create a strong signature verifier, which is the discriminator, there is a need to focus more on the quality of forgeries generated by the generator model. This work focuses on creating a generator that produces forged samples that achieve a benchmark in spoofing signature verification systems. We use CycleGANs infused with Inception model-like blocks with attention heads as the generator and a variation of the SigCNN model as the base Discriminator. We train our model with a new technique that results in 80% to 100% success in signature spoofing. Additionally, we create a custom evaluation technique to act as a goodness measure of the generated forgeries. Our work advocates generator-focused GAN architectures for spoofing data quality that aid in a better understanding of biometric data generation and evaluation.

Quantum theory is notoriously counterintuitive, and yet remains entirely self-consistent when applied universally. Here we uncover a new manifestation of its unusual consequences. We demonstrate, theoretically and experimentally (by means of polarization-encoded single-photon qubits), that Heisenberg's uncertainty principle leads to the impossibility of stringing together logical deductions about outcomes of consecutive non-compatible measurements. This phenomenon resembles the geometry of a Penrose triangle, where each corner is locally consistent while the global structure is impossible. Besides this, we show how overlooking this non-trivial logical structure leads to the erroneous possibility of distinguishing non-orthogonal states with a single measurement.

We provide a unifying framework for studying variants of topological Hochschild homology such as Real topological Hochschild homology. Associated to a crossed simplicial group $\Delta G$, a category that generalizes Connes' cyclic category, we introduce an invariant of rings with twisted $G$-action, which we call topological $\Delta G$-homology. Here a ring with twisted $G$-action is a common generalization of a ring with involution and a ring with $G$-action. Our construction provides homotopical analogues of quaternionic homology, symmetric homology, braid homology, and hyperoctahedral homology defined by Fiedorowicz--Loday. In the quaternionic case our construction, called quaternionic topological Hochschild homology, is equipped with a left Pin(2)-action and we compute it for loop spaces with a twisted $C_4$-action. As an important step along the way, we introduce a new family of crossed simplicial groups called the twisted symmetric crossed simplicial groups and we prove that they corresponds to operads for algebras with twisted $G$-action., Comment: 61 pages. Comments welcome!

Ontological commitment, i.e., used concepts, relations, and assumptions, are a corner stone of qualitative reasoning (QR) models. The state-of-the-art for processing raw inputs, though, are deep neural networks (DNNs), nowadays often based off from multimodal foundation models. These automatically learn rich representations of concepts and respective reasoning. Unfortunately, the learned qualitative knowledge is opaque, preventing easy inspection, validation, or adaptation against available QR models. So far, it is possible to associate pre-defined concepts with latent representations of DNNs, but extractable relations are mostly limited to semantic similarity. As a next step towards QR for validation and verification of DNNs: Concretely, we propose a method that extracts the learned superclass hierarchy from a multimodal DNN for a given set of leaf concepts. Under the hood we (1) obtain leaf concept embeddings using the DNN's textual input modality; (2) apply hierarchical clustering to them, using that DNNs encode semantic similarities via vector distances; and (3) label the such-obtained parent concepts using search in available ontologies from QR. An initial evaluation study shows that meaningful ontological class hierarchies can be extracted from state-of-the-art foundation models. Furthermore, we demonstrate how to validate and verify a DNN's learned representations against given ontologies. Lastly, we discuss potential future applications in the context of QR., Comment: Qualitative Reasoning Workshop 2024 (QR2024) colocated with ECAI2024, camera-ready submission; first two authors contributed equally; 10 pages, 4 figures, 3 tables

Large Language Models (LLMs) have gained popularity in task planning for long-horizon manipulation tasks. To enhance the validity of LLM-generated plans, visual demonstrations and online videos have been widely employed to guide the planning process. However, for manipulation tasks involving subtle movements but rich contact interactions, visual perception alone may be insufficient for the LLM to fully interpret the demonstration. Additionally, visual data provides limited information on force-related parameters and conditions, which are crucial for effective execution on real robots. In this paper, we introduce an in-context learning framework that incorporates tactile and force-torque information from human demonstrations to enhance LLMs' ability to generate plans for new task scenarios. We propose a bootstrapped reasoning pipeline that sequentially integrates each modality into a comprehensive task plan. This task plan is then used as a reference for planning in new task configurations. Real-world experiments on two different sequential manipulation tasks demonstrate the effectiveness of our framework in improving LLMs' understanding of multi-modal demonstrations and enhancing the overall planning performance.

Assembly is a crucial skill for robots in both modern manufacturing and service robotics. However, mastering transferable insertion skills that can handle a variety of high-precision assembly tasks remains a significant challenge. This paper presents a novel framework that utilizes diffusion models to generate 6D wrench for high-precision tactile robotic insertion tasks. It learns from demonstrations performed on a single task and achieves a zero-shot transfer success rate of 95.7% across various novel high-precision tasks. Our method effectively inherits the self-adaptability demonstrated by our previous work. In this framework, we address the frequency misalignment between the diffusion policy and the real-time control loop with a dynamic system-based filter, significantly improving the task success rate by 9.15%. Furthermore, we provide a practical guideline regarding the trade-off between diffusion models' inference ability and speed., Comment: 7 pages

Deep Neural Networks (DNNs) have become central for the perception functions of autonomous vehicles, substantially enhancing their ability to understand and interpret the environment. However, these systems exhibit inherent limitations such as brittleness, opacity, and unpredictable behavior in out-of-distribution scenarios. The European Union (EU) Artificial Intelligence (AI) Act, as a pioneering legislative framework, aims to address these challenges by establishing stringent norms and standards for AI systems, including those used in autonomous driving (AD), which are categorized as high-risk AI. In this work, we explore how the newly available generative AI models can potentially support addressing upcoming regulatory requirements in AD perception, particularly with respect to safety. This short review paper summarizes the requirements arising from the EU AI Act regarding DNN-based perception systems and systematically categorizes existing generative AI applications in AD. While generative AI models show promise in addressing some of the EU AI Acts requirements, such as transparency and robustness, this review examines their potential benefits and discusses how developers could leverage these methods to enhance compliance with the Act. The paper also highlights areas where further research is needed to ensure reliable and safe integration of these technologies.

Accurately detecting 3D objects from monocular images in dynamic roadside scenarios remains a challenging problem due to varying camera perspectives and unpredictable scene conditions. This paper introduces a two-stage training strategy to address these challenges. Our approach initially trains a model on the large-scale synthetic dataset, RoadSense3D, which offers a diverse range of scenarios for robust feature learning. Subsequently, we fine-tune the model on a combination of real-world datasets to enhance its adaptability to practical conditions. Experimental results of the Cube R-CNN model on challenging public benchmarks show a remarkable improvement in detection performance, with a mean average precision rising from 0.26 to 12.76 on the TUM Traffic A9 Highway dataset and from 2.09 to 6.60 on the DAIR-V2X-I dataset when performing transfer learning. Code, data, and qualitative video results are available on the project website: https://roadsense3d.github.io., Comment: 18 pages. Accepted for ECVA European Conference on Computer Vision 2024 (ECCV'24)

Reinforcement Learning (RL) is a potent tool for sequential decision-making and has achieved performance surpassing human capabilities across many challenging real-world tasks. As the extension of RL in the multi-agent system domain, multi-agent RL (MARL) not only need to learn the control policy but also requires consideration regarding interactions with all other agents in the environment, mutual influences among different system components, and the distribution of computational resources. This augments the complexity of algorithmic design and poses higher requirements on computational resources. Simultaneously, simulators are crucial to obtain realistic data, which is the fundamentals of RL. In this paper, we first propose a series of metrics of simulators and summarize the features of existing benchmarks. Second, to ease comprehension, we recall the foundational knowledge and then synthesize the recently advanced studies of MARL-related autonomous driving and intelligent transportation systems. Specifically, we examine their environmental modeling, state representation, perception units, and algorithm design. Conclusively, we discuss open challenges as well as prospects and opportunities. We hope this paper can help the researchers integrate MARL technologies and trigger more insightful ideas toward the intelligent and autonomous driving., Comment: 23 pages, 6 figures and 2 tables. Submitted to IEEE Journal

Arguments are presented that the reaction products of central high energy nuclear collisions up to collider energies can rigorously be interpreted in terms of a continuous decoupling mechanism based on continuous equations of motion. The various aspects of the collision dynamics are investigate in terms of the individual decoupling processes. Thereby each observed particle decouples during its own temporal decoupling window. This includes a ``very late'' decoupling of the faintly bound Hypertritons observed in recent ALICE experiments. The success of the strategy is based upon 200 years old wisdom and leads to a revised interpretation of the entire decoupling process., Comment: 6 pages, 3 figures, submitted to Physics Letters B

Existing roadside perception systems are limited by the absence of publicly available, large-scale, high-quality 3D datasets. Exploring the use of cost-effective, extensive synthetic datasets offers a viable solution to tackle this challenge and enhance the performance of roadside monocular 3D detection. In this study, we introduce the TUMTraf Synthetic Dataset, offering a diverse and substantial collection of high-quality 3D data to augment scarce real-world datasets. Besides, we present WARM-3D, a concise yet effective framework to aid the Sim2Real domain transfer for roadside monocular 3D detection. Our method leverages cheap synthetic datasets and 2D labels from an off-the-shelf 2D detector for weak supervision. We show that WARM-3D significantly enhances performance, achieving a +12.40% increase in mAP 3D over the baseline with only pseudo-2D supervision. With 2D GT as weak labels, WARM-3D even reaches performance close to the Oracle baseline. Moreover, WARM-3D improves the ability of 3D detectors to unseen sample recognition across various real-world environments, highlighting its potential for practical applications.

We introduce DexGanGrasp, a dexterous grasping synthesis method that generates and evaluates grasps with single view in real time. DexGanGrasp comprises a Conditional Generative Adversarial Networks (cGANs)-based DexGenerator to generate dexterous grasps and a discriminator-like DexEvalautor to assess the stability of these grasps. Extensive simulation and real-world expriments showcases the effectiveness of our proposed method, outperforming the baseline FFHNet with an 18.57% higher success rate in real-world evaluation. We further extend DexGanGrasp to DexAfford-Prompt, an open-vocabulary affordance grounding pipeline for dexterous grasping leveraging Multimodal Large Language Models (MLLMs) and Vision Language Models (VLMs), to achieve task-oriented grasping with successful real-world deployments., Comment: 8 pages, 4 figures

Synthesizing diverse dexterous grasps from uncertain partial observation is an important yet challenging task for physically intelligent embodiments. Previous works on generative grasp synthesis fell short of precisely capturing the complex grasp distribution and reasoning about shape uncertainty in the unstructured and often partially perceived reality. In this work, we introduce a novel model that can generate diverse grasps for a multi-fingered hand while introspectively handling perceptual uncertainty and recognizing unknown object geometry to avoid performance degradation. Specifically, we devise a Deep Latent Variable Model (DLVM) based on Normalizing Flows (NFs), facilitating hierarchical and expressive latent representation for modeling versatile grasps. Our model design counteracts typical pitfalls of its popular alternative in generative grasping, i.e., conditional Variational Autoencoders (cVAEs) whose performance is limited by mode collapse and miss-specified prior issues. Moreover, the resultant feature hierarchy and the exact flow likelihood computation endow our model with shape-aware introspective capabilities, enabling it to quantify the shape uncertainty of partial point clouds and detect objects of novel geometry. We further achieve performance gain by fusing this information with a discriminative grasp evaluator, facilitating a novel hybrid way for grasp evaluation. Comprehensive simulated and real-world experiments show that the proposed idea gains superior performance and higher run-time efficiency against strong baselines, including diffusion models. We also demonstrate substantial benefits of greater diversity for grasping objects in clutter and a confined workspace in the real world., Comment: First two authors contributed equally, whose ordering decided via coin-tossing. Under Reivew

In frame-based vision, object detection faces substantial performance degradation under challenging conditions due to the limited sensing capability of conventional cameras. Event cameras output sparse and asynchronous events, providing a potential solution to solve these problems. However, effectively fusing two heterogeneous modalities remains an open issue. In this work, we propose a novel hierarchical feature refinement network for event-frame fusion. The core concept is the design of the coarse-to-fine fusion module, denoted as the cross-modality adaptive feature refinement (CAFR) module. In the initial phase, the bidirectional cross-modality interaction (BCI) part facilitates information bridging from two distinct sources. Subsequently, the features are further refined by aligning the channel-level mean and variance in the two-fold adaptive feature refinement (TAFR) part. We conducted extensive experiments on two benchmarks: the low-resolution PKU-DDD17-Car dataset and the high-resolution DSEC dataset. Experimental results show that our method surpasses the state-of-the-art by an impressive margin of $\textbf{8.0}\%$ on the DSEC dataset. Besides, our method exhibits significantly better robustness (\textbf{69.5}\% versus \textbf{38.7}\%) when introducing 15 different corruption types to the frame images. The code can be found at the link (https://github.com/HuCaoFighting/FRN)., Comment: Accepted by ECCV 2024

3D point cloud segmentation has received significant interest for its growing applications. However, the generalization ability of models suffers in dynamic scenarios due to the distribution shift between test and training data. To promote robustness and adaptability across diverse scenarios, test-time adaptation (TTA) has recently been introduced. Nevertheless, most existing TTA methods are developed for images, and limited approaches applicable to point clouds ignore the inherent hierarchical geometric structures in point cloud streams, i.e., local (point-level), global (object-level), and temporal (frame-level) structures. In this paper, we delve into TTA in 3D point cloud segmentation and propose a novel Hierarchical Geometry Learning (HGL) framework. HGL comprises three complementary modules from local, global to temporal learning in a bottom-up manner.Technically, we first construct a local geometry learning module for pseudo-label generation. Next, we build prototypes from the global geometry perspective for pseudo-label fine-tuning. Furthermore, we introduce a temporal consistency regularization module to mitigate negative transfer. Extensive experiments on four datasets demonstrate the effectiveness and superiority of our HGL. Remarkably, on the SynLiDAR to SemanticKITTI task, HGL achieves an overall mIoU of 46.91\%, improving GIPSO by 3.0\% and significantly reducing the required adaptation time by 80\%. The code is available at https://github.com/tpzou/HGL.

Transformer-based networks applied to image patches have achieved cutting-edge performance in many vision tasks. However, lacking the built-in bias of convolutional neural networks (CNN) for local image statistics, they require large datasets and modifications to capture relationships between patches, especially in segmentation tasks. Images in the frequency domain might be more suitable for the attention mechanism, as local features are represented globally. By transforming images into the frequency domain, local features are represented globally. Due to MRI data acquisition properties, these images are particularly suitable. This work investigates how the image domain (spatial or k-space) affects segmentation results of deep learning (DL) models, focusing on attention-based networks and other non-convolutional models based on MLPs. We also examine the necessity of additional positional encoding for Transformer-based networks when input images are in the frequency domain. For evaluation, we pose a skull stripping task and a brain tissue segmentation task. The attention-based models used are PerceiverIO and a vanilla Transformer encoder. To compare with non-attention-based models, an MLP and ResMLP are also trained and tested. Results are compared with the Swin-Unet, the state-of-the-art medical image segmentation model. Experimental results show that using k-space for the input domain can significantly improve segmentation results. Also, additional positional encoding does not seem beneficial for attention-based networks if the input is in the frequency domain. Although none of the models matched the Swin-Unet's performance, the less complex models showed promising improvements with a different domain choice., Comment: 13 pages, 2 figures

Learning from demonstrations faces challenges in generalizing beyond the training data and is fragile even to slight visual variations. To tackle this problem, we introduce Lan-o3dp, a language guided object centric diffusion policy that takes 3d representation of task relevant objects as conditional input and can be guided by cost function for safety constraints at inference time. Lan-o3dp enables strong generalization in various aspects, such as background changes, visual ambiguity and can avoid novel obstacles that are unseen during the demonstration process. Specifically, We first train a diffusion policy conditioned on point clouds of target objects and then harness a large language model to decompose the user instruction into task related units consisting of target objects and obstacles, which can be used as visual observation for the policy network or converted to a cost function, guiding the generation of trajectory towards collision free region at test time. Our proposed method shows training efficiency and higher success rates compared with the baselines in simulation experiments. In real world experiments, our method exhibits strong generalization performance towards unseen instances, cluttered scenes, scenes of multiple similar objects and demonstrates training free capability of obstacle avoidance., Comment: 11 pages, 8 figures

3D Gaussian Splatting (3DGS) has emerged as a cutting-edge technique for real-time radiance field rendering, offering state-of-the-art performance in terms of both quality and speed. 3DGS models a scene as a collection of three-dimensional Gaussians, or splats, with additional attributes optimized to conform to the scene's geometric and visual properties. Despite its advantages in rendering speed and image fidelity, 3DGS is limited by its significant storage and memory demands. These high demands make 3DGS impractical for mobile devices or headsets, reducing its applicability in important areas of computer graphics. To address these challenges and advance the practicality of 3DGS, this survey provides a comprehensive and detailed examination of compression and compaction techniques developed to make 3DGS more efficient. We categorize current approaches into compression techniques, which aim at achieving the highest quality at minimal data size, and compaction techniques, which aim for optimal quality with the fewest Gaussians. We introduce the basic mathematical concepts underlying the analyzed methods, as well as key implementation details and design choices. Our report thoroughly discusses similarities and differences among the methods, as well as their respective advantages and disadvantages. We establish a consistent standard for comparing these methods based on key performance metrics and datasets. Specifically, since these methods have been developed in parallel and over a short period of time, currently, no comprehensive comparison exists. This survey, for the first time, presents a unified standard to evaluate 3DGS compression techniques. To facilitate the continuous monitoring of emerging methodologies, we maintain a dedicated website that will be regularly updated with new techniques and revisions of existing findings https://w-m.github.io/3dgs-compression-survey/ ., Comment: 3D Gaussian Splatting compression survey; 3DGS compression; new approaches added

This data curation work introduces the first large-scale dataset of radial k-space and DICOM data for breast DCE-MRI acquired in diagnostic breast MRI exams. Our dataset includes case-level labels indicating patient age, menopause status, lesion status (negative, benign, and malignant), and lesion type for each case. The public availability of this dataset and accompanying reconstruction code will support research and development of fast and quantitative breast image reconstruction and machine learning methods.

As transformer-based object detection models progress, their impact in critical sectors like autonomous vehicles and aviation is expected to grow. Soft errors causing bit flips during inference have significantly impacted DNN performance, altering predictions. Traditional range restriction solutions for CNNs fall short for transformers. This study introduces the Global Clipper and Global Hybrid Clipper, effective mitigation strategies specifically designed for transformer-based models. It significantly enhances their resilience to soft errors and reduces faulty inferences to ~ 0\%. We also detail extensive testing across over 64 scenarios involving two transformer models (DINO-DETR and Lite-DETR) and two CNN models (YOLOv3 and SSD) using three datasets, totalling approximately 3.3 million inferences, to assess model robustness comprehensively. Moreover, the paper explores unique aspects of attention blocks in transformers and their operational differences from CNNs., Comment: Accepted at IJCAI-AISafety'24 Workshop

We introduce Situation Monitor, a novel zero-shot Out-of-Distribution (OOD) detection approach for transformer-based object detection models to enhance reliability in safety-critical machine learning applications such as autonomous driving. The Situation Monitor utilizes the Diversity-based Budding Ensemble Architecture (DBEA) and increases the OOD performance by integrating a diversity loss into the training process on top of the budding ensemble architecture, detecting Far-OOD samples and minimizing false positives on Near-OOD samples. Moreover, utilizing the resulting DBEA increases the model's OOD performance and improves the calibration of confidence scores, particularly concerning the intersection over union of the detected objects. The DBEA model achieves these advancements with a 14% reduction in trainable parameters compared to the vanilla model. This signifies a substantial improvement in efficiency without compromising the model's ability to detect OOD instances and calibrate the confidence scores accurately., Comment: Paper accepted at CVPR SAIAD Workshop

Recently, Large Language Models (LLMs) have witnessed remarkable performance as zero-shot task planners for robotic manipulation tasks. However, the open-loop nature of previous works makes LLM-based planning error-prone and fragile. On the other hand, failure detection approaches for closed-loop planning are often limited by task-specific heuristics or following an unrealistic assumption that the prediction is trustworthy all the time. As a general-purpose reasoning machine, LLMs or Multimodal Large Language Models (MLLMs) are promising for detecting failures. However, However, the appropriateness of the aforementioned assumption diminishes due to the notorious hullucination problem. In this work, we attempt to mitigate these issues by introducing a framework for closed-loop LLM-based planning called KnowLoop, backed by an uncertainty-based MLLMs failure detector, which is agnostic to any used MLLMs or LLMs. Specifically, we evaluate three different ways for quantifying the uncertainty of MLLMs, namely token probability, entropy, and self-explained confidence as primary metrics based on three carefully designed representative prompting strategies. With a self-collected dataset including various manipulation tasks and an LLM-based robot system, our experiments demonstrate that token probability and entropy are more reflective compared to self-explained confidence. By setting an appropriate threshold to filter out uncertain predictions and seek human help actively, the accuracy of failure detection can be significantly enhanced. This improvement boosts the effectiveness of closed-loop planning and the overall success rate of tasks., Comment: Accepted at ICRA 2024 Workshop on Back to the Future: Robot Learning Going Probabilistic. Website: https://sites.google.com/view/konwloop/home

Safe reinforcement learning (RL) is crucial for deploying RL agents in real-world applications, as it aims to maximize long-term rewards while satisfying safety constraints. However, safe RL often suffers from sample inefficiency, requiring extensive interactions with the environment to learn a safe policy. We propose Efficient Safe Policy Optimization (ESPO), a novel approach that enhances the efficiency of safe RL through sample manipulation. ESPO employs an optimization framework with three modes: maximizing rewards, minimizing costs, and balancing the trade-off between the two. By dynamically adjusting the sampling process based on the observed conflict between reward and safety gradients, ESPO theoretically guarantees convergence, optimization stability, and improved sample complexity bounds. Experiments on the Safety-MuJoCo and Omnisafe benchmarks demonstrate that ESPO significantly outperforms existing primal-based and primal-dual-based baselines in terms of reward maximization and constraint satisfaction. Moreover, ESPO achieves substantial gains in sample efficiency, requiring 25--29% fewer samples than baselines, and reduces training time by 21--38%.

In numerous reinforcement learning (RL) problems involving safety-critical systems, a key challenge lies in balancing multiple objectives while simultaneously meeting all stringent safety constraints. To tackle this issue, we propose a primal-based framework that orchestrates policy optimization between multi-objective learning and constraint adherence. Our method employs a novel natural policy gradient manipulation method to optimize multiple RL objectives and overcome conflicting gradients between different tasks, since the simple weighted average gradient direction may not be beneficial for specific tasks' performance due to misaligned gradients of different task objectives. When there is a violation of a hard constraint, our algorithm steps in to rectify the policy to minimize this violation. We establish theoretical convergence and constraint violation guarantees in a tabular setting. Empirically, our proposed method also outperforms prior state-of-the-art methods on challenging safe multi-objective reinforcement learning tasks.

The Bethe free energy approximation provides an effective way for relaxing NP-hard problems of probabilistic inference. However, its accuracy depends on the model parameters and particularly degrades if a phase transition in the model occurs. In this work, we analyze when the Bethe approximation is reliable and how this can be verified. We argue and show by experiment that it is mostly accurate if it is convex on a submanifold of its domain, the 'Bethe box'. For verifying its convexity, we derive two sufficient conditions that are based on the definiteness properties of the Bethe Hessian matrix: the first uses the concept of diagonal dominance, and the second decomposes the Bethe Hessian matrix into a sum of sparse matrices and characterizes the definiteness properties of the individual matrices in that sum. These theoretical results provide a simple way to estimate the critical phase transition temperature of a model. As a practical contribution we propose $\texttt{BETHE-MIN}$, a projected quasi-Newton method to efficiently find a minimum of the Bethe free energy., Comment: This work has been submitted to the Journal of Machine Learning Research (JMLR) for possible publication