Search

Trajectory prediction forecasts nearby agents' moves based on their historical trajectories. Accurate trajectory prediction is crucial for autonomous vehicles. Existing attacks compromise the prediction model of a victim AV by directly manipulating the historical trajectory of an attacker AV, which has limited real-world applicability. This paper, for the first time, explores an indirect attack approach that induces prediction errors via attacks against the perception module of a victim AV. Although it has been shown that physically realizable attacks against LiDAR-based perception are possible by placing a few objects at strategic locations, it is still an open challenge to find an object location from the vast search space in order to launch effective attacks against prediction under varying victim AV velocities. Through analysis, we observe that a prediction model is prone to an attack focusing on a single point in the scene. Consequently, we propose a novel two-stage attack framework to realize the single-point attack. The first stage of prediction-side attack efficiently identifies, guided by the distribution of detection results under object-based attacks against perception, the state perturbations for the prediction model that are effective and velocity-insensitive. In the second stage of location matching, we match the feasible object locations with the found state perturbations. Our evaluation using a public autonomous driving dataset shows that our attack causes a collision rate of up to 63% and various hazardous responses of the victim AV. The effectiveness of our attack is also demonstrated on a real testbed car. To the best of our knowledge, this study is the first security analysis spanning from LiDAR-based perception to prediction in autonomous driving, leading to a realistic attack on prediction. To counteract the proposed attack, potential defenses are discussed., Comment: In Proceedings of the 33rd USENIX Security Symposium 2024

Horizontally aligned carbon nanotube (HACNT) arrays hold significant potential for various applications in nanoelectronics and material science. However, their high-throughput characterization remains challenging due to the lack of methods with both high efficiency and high accuracy. Here, we present a novel technique, Calibrated Absolute Optical Contrast (CAOC), achieved through the implementation of differential principles to filter out stray signals and high-resolution calibration to endow optical contrast with physical significance. CAOC offers major advantages over previous characterization techniques, providing consistent and reliable measurements of HACNT array density with high throughput and non-destructive assessment. To validate its utility, we demonstrate wafer-scale uniformity assessment by rapid density mapping. This technique not only facilitates the practical evaluation of HACNT arrays but also provides insights into balancing high throughput and high resolution in nanomaterial characterization.

Simulating realistic interactions among traffic agents is crucial for efficiently validating the safety of autonomous driving systems. Existing leading simulators primarily use an encoder-decoder structure to encode the historical trajectories for future simulation. However, such a paradigm complicates the model architecture, and the manual separation of history and future trajectories leads to low data utilization. To address these challenges, we propose Behavior Generative Pre-trained Transformers (BehaviorGPT), a decoder-only, autoregressive architecture designed to simulate the sequential motion of multiple agents. Crucially, our approach discards the traditional separation between "history" and "future," treating each time step as the "current" one, resulting in a simpler, more parameter- and data-efficient design that scales seamlessly with data and computation. Additionally, we introduce the Next-Patch Prediction Paradigm (NP3), which enables models to reason at the patch level of trajectories and capture long-range spatial-temporal interactions. BehaviorGPT ranks first across several metrics on the Waymo Sim Agents Benchmark, demonstrating its exceptional performance in multi-agent and agent-map interactions. We outperformed state-of-the-art models with a realism score of 0.741 and improved the minADE metric to 1.540, with an approximately 91.6% reduction in model parameters.

In this paper, we prove a nonabelian Hodge correspondence for principal bundles on a smooth variety $X$ in positive characteristic, which generalizes the Ogus-Vologodsky correspondence for vector bundles. Then we extend the correspondence to logahoric torsors over a log pair $(X,D)$, where $D$ a reduced normal crossing divisor in $X$. As an intermediate step, we prove a correspondence between principal bundles on root stacks $\mathscr{X}$ and parahoric torsors on $(X,D)$, which generalizes the correspondence on curves given by Balaji--Seshadri to higher dimensional case., Comment: 29 pages

In this paper, we study the existence of gr-semistable filtrations of orthogonal/symplectic $\lambda$-connections. It is known that gr-semistable filtrations always exist for flat bundles in arbitrary characteristic. However, we found a counterexample of orthogonal flat bundles of rank 5 in positive characteristic. The central new idea in this example is the notion of quasi gr-semistability for orthogonal/symplectic $\lambda$-connections. We establish the equivalence between gr-semistability and quasi gr-semistablity for an orthogonal/symplectic $\lambda$-connection. This provides a way to determine whether an orthogonal/symplectic $\lambda$-connection is gr-semistable. As an application, we obtain a characterization of gr-semistable orthogonal $\lambda$-connections of rank $\leq 6$., Comment: 35 pages

Modern database management systems (DBMS) expose hundreds of configurable knobs to control system behaviours. Determining the appropriate values for these knobs to improve DBMS performance is a long-standing problem in the database community. As there is an increasing number of knobs to tune and each knob could be in continuous or categorical values, manual tuning becomes impractical. Recently, automatic tuning systems using machine learning methods have shown great potentials. However, existing approaches still incur significant tuning costs or only yield sub-optimal performance. This is because they either ignore the extensive domain knowledge available (e.g., DBMS manuals and forum discussions) and only rely on the runtime feedback of benchmark evaluations to guide the optimization, or they utilize the domain knowledge in a limited way. Hence, we propose GPTuner, a manual-reading database tuning system. Firstly, we develop a Large Language Model (LLM)-based pipeline to collect and refine heterogeneous knowledge, and propose a prompt ensemble algorithm to unify a structured view of the refined knowledge. Secondly, using the structured knowledge, we (1) design a workload-aware and training-free knob selection strategy, (2) develop a search space optimization technique considering the value range of each knob, and (3) propose a Coarse-to-Fine Bayesian Optimization Framework to explore the optimized space. Finally, we evaluate GPTuner under different benchmarks (TPC-C and TPC-H), metrics (throughput and latency) as well as DBMS (PostgreSQL and MySQL). Compared to the state-of-the-art approaches, GPTuner identifies better configurations in 16x less time on average. Moreover, GPTuner achieves up to 30% performance improvement (higher throughput or lower latency) over the best-performing alternative., Comment: 15 pages, 14 figures, submit to VLDB2024

Scanning electron microscopy (SEM) is indispensable in diverse applications ranging from microelectronics to food processing because it provides large depth-of-field images with a resolution beyond the optical diffraction limit. However, the technology requires coating conductive films on insulator samples and a vacuum environment. We use deep learning to obtain the mapping relationship between optical super-resolution (OSR) images and SEM domain images, which enables the transformation of OSR images into SEM-like large depth-of-field images. Our custom-built scanning superlens microscopy (SSUM) system, which requires neither coating samples by conductive films nor a vacuum environment, is used to acquire the OSR images with features down to ~80 nm. The peak signal-to-noise ratio (PSNR) and structural similarity index measure values indicate that the deep learning method performs excellently in image-to-image translation, with a PSNR improvement of about 0.74 dB over the optical super-resolution images. The proposed method provides a high level of detail in the reconstructed results, indicating that it has broad applicability to chip-level defect detection, biological sample analysis, forensics, and various other fields., Comment: 13 pages,7 figures

Numerical simulations were conducted to investigate the flow field characteristics and performance of a carbon-hydrogen/oxygen-rich air rotating detonation engine (RDE). Three distinct flow field structures were observed in the gas-solid two-phase RDE. The results show that reducing the hydrogen equivalence ratio and particle diameter both contribute to the transition from gas-phase single-front detonation to gas-solid two-phase double-front detonation and further to gas-solid two-phase single-front detonation. The effects of solid fuel particle diameter and hydrogen equivalence ratio on the flow field characteristics and performance are revealed. The results show that reducing the particle diameter enhances the speed of the two-phase detonation wave, improves the pressure gain in the combustion chamber, and increases the specific impulse. Decreasing the hydrogen equivalence ratio reduces the detonation wave speed, enhances the stability of the detonation flow field, increases the pressure gain in the detonation wave and combustion chamber and boosts thrust. Furthermore, the selection of operational conditions to ensure stable operation and optimal performance of the RDE is discussed. In order to take into account the requirements of stability, pressure gain performance and propulsion performance, two-phase single-front detonation should be realized in gas-solid two-phase RDE, and smaller hydrogen equivalent ratio and appropriate particle diameter should be selected. According to the conclusion of this study, the particle diameter should be 0.5-1 {\mu}m. Under such conditions, the detonation flow field demonstrates good stability, allowing the RDE to achieve higher pressure gain and specific impulse while maintaining stable operation.

Multi-modal fusion has shown initial promising results for object detection of autonomous driving perception. However, many existing fusion schemes do not consider the quality of each fusion input and may suffer from adverse conditions on one or more sensors. While predictive uncertainty has been applied to characterize single-modal object detection performance at run time, incorporating uncertainties into the multi-modal fusion still lacks effective solutions due primarily to the uncertainty's cross-modal incomparability and distinct sensitivities to various adverse conditions. To fill this gap, this paper proposes Uncertainty-Encoded Mixture-of-Experts (UMoE) that explicitly incorporates single-modal uncertainties into LiDAR-camera fusion. UMoE uses individual expert network to process each sensor's detection result together with encoded uncertainty. Then, the expert networks' outputs are analyzed by a gating network to determine the fusion weights. The proposed UMoE module can be integrated into any proposal fusion pipeline. Evaluation shows that UMoE achieves a maximum of 10.67%, 3.17%, and 5.40% performance gain compared with the state-of-the-art proposal-level multi-modal object detectors under extreme weather, adversarial, and blinding attack scenarios., Comment: In proceedings of the 26th European Conference on Artificial Intelligence ECAI 2023. 8 pages + 2 appendix pages

Estimating the joint distribution of on-road agents' future trajectories is essential for autonomous driving. In this technical report, we propose a next-generation framework for joint multi-agent trajectory prediction called QCNeXt. First, we adopt the query-centric encoding paradigm for the task of joint multi-agent trajectory prediction. Powered by this encoding scheme, our scene encoder is equipped with permutation equivariance on the set elements, roto-translation invariance in the space dimension, and translation invariance in the time dimension. These invariance properties not only enable accurate multi-agent forecasting fundamentally but also empower the encoder with the capability of streaming processing. Second, we propose a multi-agent DETR-like decoder, which facilitates joint multi-agent trajectory prediction by modeling agents' interactions at future time steps. For the first time, we show that a joint prediction model can outperform marginal prediction models even on the marginal metrics, which opens up new research opportunities in trajectory prediction. Our approach ranks 1st on the Argoverse 2 multi-agent motion forecasting benchmark, winning the championship of the Argoverse Challenge at the CVPR 2023 Workshop on Autonomous Driving., Comment: Technical report for the 1st place solution of the Argoverse 2 Multi-Agent Motion Forecasting Competition at the CVPR 2023 Workshop on Autonomous Driving

In autonomous driving, an accurate understanding of environment, e.g., the vehicle-to-vehicle and vehicle-to-lane interactions, plays a critical role in many driving tasks such as trajectory prediction and motion planning. Environment information comes from high-definition (HD) map and historical trajectories of vehicles. Due to the heterogeneity of the map data and trajectory data, many data-driven models for trajectory prediction and motion planning extract vehicle-to-vehicle and vehicle-to-lane interactions in a separate and sequential manner. However, such a manner may capture biased interpretation of interactions, causing lower prediction and planning accuracy. Moreover, separate extraction leads to a complicated model structure and hence the overall efficiency and scalability are sacrificed. To address the above issues, we propose an environment representation, Temporal Occupancy Flow Graph (TOFG). Specifically, the occupancy flow-based representation unifies the map information and vehicle trajectories into a homogeneous data format and enables a consistent prediction. The temporal dependencies among vehicles can help capture the change of occupancy flow timely to further promote model performance. To demonstrate that TOFG is capable of simplifying the model architecture, we incorporate TOFG with a simple graph attention (GAT) based neural network and propose TOFG-GAT, which can be used for both trajectory prediction and motion planning. Experiment results show that TOFG-GAT achieves better or competitive performance than all the SOTA baselines with less training time., Comment: Accepted by ICRA 2023

Predicting the future trajectories of nearby objects plays a pivotal role in Robotics and Automation such as autonomous driving. While learning-based trajectory prediction methods have achieved remarkable performance on public benchmarks, the generalization ability of these approaches remains questionable. The poor generalizability on unseen domains, a well-recognized defect of data-driven approaches, can potentially harm the real-world performance of trajectory prediction models. We are thus motivated to improve generalization ability of models instead of merely pursuing high accuracy on average. Due to the lack of benchmarks for quantifying the generalization ability of trajectory predictors, we first construct a new benchmark called argoverse-shift, where the data distributions of domains are significantly different. Using this benchmark for evaluation, we identify that the domain shift problem seriously hinders the generalization of trajectory predictors since state-of-the-art approaches suffer from severe performance degradation when facing those out-of-distribution scenes. To enhance the robustness of models against domain shift problems, we propose a plug-and-play strategy for domain normalization in trajectory prediction. Our strategy utilizes the Frenet coordinate frame for modeling and can effectively narrow the domain gap of different scenes caused by the variety of road geometry and topology. Experiments show that our strategy noticeably boosts the prediction performance of the state-of-the-art in domains that were previously unseen to the models, thereby improving the generalization ability of data-driven trajectory prediction methods., Comment: This paper was published in 2023 IEEE International Conference on Robotics and Automation (ICRA). New version updated with links to the source code of the Frenet+ strategy

Data augmentation (DA) is indispensable in modern machine learning and deep neural networks. The basic idea of DA is to construct new training data to improve the model's generalization by adding slightly disturbed versions of existing data or synthesizing new data. This survey comprehensively reviews a crucial subset of DA techniques, namely Mix-based Data Augmentation (MixDA), which generates novel samples by combining multiple examples. In contrast to traditional DA approaches that operate on single samples or entire datasets, MixDA stands out due to its effectiveness, simplicity, flexibility, computational efficiency, theoretical foundation, and broad applicability. We begin by introducing a novel taxonomy that categorizes MixDA into Mixup-based, Cutmix-based, and mixture approaches based on a hierarchical perspective of the data mixing operation. Subsequently, we provide an in-depth review of various MixDA techniques, focusing on their underlying motivations. Owing to its versatility, MixDA has penetrated a wide range of applications, which we also thoroughly investigate in this survey. Moreover, we delve into the underlying mechanisms of MixDA's effectiveness by examining its impact on model generalization and calibration while providing insights into the model's behavior by analyzing the inherent properties of MixDA. Finally, we recapitulate the critical findings and fundamental challenges of current MixDA studies while outlining the potential directions for future works. Different from previous related surveys that focus on DA approaches in specific domains (e.g., CV and NLP) or only review a limited subset of MixDA studies, we are the first to provide a systematical survey of MixDA, covering its taxonomy, methodology, application, and explainability. Furthermore, we provide promising directions for researchers interested in this exciting area., Comment: 41 pages, 4 figures, and 5 tables

We establish a torsion theorem to the effect that the unique zero of the Kodaira-Spencer map attached to a certain quasi-semistable family of complex projective varieties over the complex projective line is the image of a torsion point of an elliptic curve under the natural projection. The proof is a mod $p$ argument and requires a density one set of primes. There are three essential ingredients in the proof: a solution to the conjecture of Sun-Yang-Zuo, which constitutes the principal part of the paper, Pink's theorem, and Higgs periodicity theorem., Comment: 3rd version. Acknowledgement to the initial discovery of the torsion phenomenon due to Jingbang Yang and Kang Zuo as proved in the paper added

A generalized matrix-pencil approach is proposed for the estimation of complex exponential components with segmented signal samples, which is very efficient and provides super-resolution estimations. It is applicable to the signals sampled segmentally with the same sampling frequency and direction of arrival (DOA) estimation with distributed arrays within which array elements are placed uniformly with the same inter-element spacing., Comment: 5 pages, 6 Figures

Car-following (CF) modeling, an essential component in simulating human CF behaviors, has attracted increasing research interest in the past decades. This paper pushes the state of the art by proposing a novel generative hybrid CF model, which achieves high accuracy in characterizing dynamic human CF behaviors and is able to generate realistic human CF behaviors for any given observed or even unobserved driving style. Specifically, the ability of accurately capturing human CF behaviors is ensured by designing and calibrating an Intelligent Driver Model (IDM) with time-varying parameters. The reason behind is that such time-varying parameters can express both the inter-driver heterogeneity, i.e., diverse driving styles of different drivers, and the intra-driver heterogeneity, i.e., changing driving styles of the same driver. The ability of generating realistic human CF behaviors of any given observed driving style is achieved by applying a neural process (NP) based model. The ability of inferring CF behaviors of unobserved driving styles is supported by exploring the relationship between the calibrated time-varying IDM parameters and an intermediate variable of NP. To demonstrate the effectiveness of our proposed models, we conduct extensive experiments and comparisons, including CF model parameter calibration, CF behavior prediction, and trajectory simulation for different driving styles.

Age-of-Information (AoI) is an application layer metric that has been widely adopted to quantify the information freshness of each information source. However, few works address the impact of probabilistic transmission failures on satisfying the harsh AoI requirement of each source, which is of critical importance in a great number of wireless-powered real-time applications. In this paper, we investigate the transmission scheduling problem of maximizing throughput over wireless channels under different time-average AoI requirements for heterogeneous information sources. When the channel reliability for each source is known as prior, the global optimal transmission scheduling policy is proposed. Moreover, when channel reliabilities are unknown, it is modeled as an AoI-constrained Multi-Armed Bandit (MAB) problem. Then a learning algorithm that meets the AoI requirement with probability 1 and incurs up to O(K\sqrt{T\log T}) accumulated regret is proposed, where K is the number of arms/information sources, and T is the time horizon. Numerical results show that the accumulated regret of our learning algorithm is strictly bounded by K\sqrt{T\log T} and outperforms the AoI-constraint-aware baseline, and the AoI requirement of every source is robustly satisfied., Comment: technical errors are found in the proof