Searcharxiv⌕ Search

arXiv subjects

Jun Ma

Publications and source records attributed to Jun Ma.

At least 271 records · Page 15Linked to original sources
arXiv2023

Integrated Behavior Planning and Motion Control for Autonomous Vehicles with Traffic Rules Compliance

In this article, we propose an optimization-based integrated behavior planning and motion control scheme, which is an interpretable and adaptable urban autonomous driving solution that complies with complex traffic rules while ensuring driving safety. Inherently, to ensure compliance with traffic rules, an innovative design of potential functions (PFs) is presented to characterize various traffic rules related to traffic lights, traversable and non-traversable traffic line markings, etc. These PFs are further incorporated as part of the model predictive control (MPC) formulation. In this sense, high-level behavior planning is attained implicitly along with motion control as an integrated architecture, facilitating flexible maneuvers with safety guarantees. Due to the well-designed objective function of the MPC scheme, our integrated behavior planning and motion control scheme is competent for various urban driving scenarios and able to generate versatile behaviors, such as overtaking with adaptive cruise control, turning in the intersection, and merging in and out of the roundabout. As demonstrated from a series of simulations with challenging scenarios in CARLA, it is noteworthy that the proposed framework admits real-time performance and high generalizability.

cs.RO↗
arXiv2023

GPT-4V(ision) Unsuitable for Clinical Care and Education: A Clinician-Evaluated Assessment

OpenAI's large multimodal model, GPT-4V(ision), was recently developed for general image interpretation. However, less is known about its capabilities with medical image interpretation and diagnosis. Board-certified physicians and senior residents assessed GPT-4V's proficiency across a range of medical conditions using imaging modalities such as CT scans, MRIs, ECGs, and clinical photographs. Although GPT-4V is able to identify and explain medical images, its diagnostic accuracy and clinical decision-making abilities are poor, posing risks to patient safety. Despite the potential that large language models may have in enhancing medical education and delivery, the current limitations of GPT-4V in interpreting medical images reinforces the importance of appropriate caution when using it for clinical decision-making.

cs.CV↗
arXiv2023

Learning Robust Sequential Recommenders through Confident Soft Labels

Sequential recommenders that are trained on implicit feedback are usually learned as a multi-class classification task through softmax-based loss functions on one-hot class labels. However, one-hot training labels are sparse and may lead to biased training and sub-optimal performance. Dense, soft labels have been shown to help improve recommendation performance. But how to generate high-quality and confident soft labels from noisy sequential interactions between users and items is still an open question. We propose a new learning framework for sequential recommenders, CSRec, which introduces confident soft labels to provide robust guidance when learning from user-item interactions. CSRec contains a teacher module that generates high-quality and confident soft labels and a student module that acts as the target recommender and is trained on the combination of dense, soft labels and sparse, one-hot labels. We propose and compare three approaches to constructing the teacher module: (i) model-level, (ii) data-level, and (iii) training-level. To evaluate the effectiveness and generalization ability of CSRec, we conduct experiments using various state-of-the-art sequential recommendation models as the target student module on four benchmark datasets. Our experimental results demonstrate that CSRec is effective in training better performing sequential recommenders.

cs.IR↗
arXiv2023

The Qitai Radio Telescope

This study presents a general outline of the Qitai radio telescope (QTT) project. Qitai, the site of the telescope, is a county of Xinjiang Uygur Autonomous Region of China, located in the east Tianshan Mountains at an elevation of about 1800 m. The QTT is a fully steerable, Gregorian type telescope with a standard parabolic main reflector of 110 m diameter. The QTT has adopted an um-brella support, homology-symmetric lightweight design. The main reflector is active so that the deformation caused by gravity can be corrected. The structural design aims to ultimately allow high-sensitivity observations from 150 MHz up to 115 GHz. To satisfy the requirements for early scientific goals, the QTT will be equipped with ultra-wideband receivers and large field-of-view mul-ti-beam receivers. A multi-function signal-processing system based on RFSoC and GPU processor chips will be developed. These will enable the QTT to operate in pulsar, spectral line, continuum and Very Long Baseline Interferometer (VLBI) observing modes. Electromagnetic compatibility (EMC) and radio frequency interference (RFI) control techniques are adopted throughout the system design. The QTT will form a world-class observational platform for the detection of low-frequency (nanoHertz) gravitational waves through pulsar timing array (PTA) techniques, pulsar surveys, the discovery of binary black-hole systems, and exploring dark matter and the origin of life in the universe.

astro-ph.IM↗
arXiv2023

Geometry-Aware Safety-Critical Local Reactive Controller for Robot Navigation in Unknown and Cluttered Environments

This work proposes a safety-critical local reactive controller that enables the robot to navigate in unknown and cluttered environments. In particular, the trajectory tracking task is formulated as a constrained polynomial optimization problem. Then, safety constraints are imposed on the control variables invoking the notion of polynomial positivity certificates in conjunction with their Sum-of-Squares (SOS) approximation, thereby confining the robot motion inside the locally extracted convex free region. It is noteworthy that, in the process of devising the proposed safety constraints, the geometry of the robot can be approximated using any shape that can be characterized with a set of polynomial functions. The optimization problem is further convexified into a semidefinite program (SDP) leveraging truncated multi-sequences (tms) and moment relaxation, which favorably facilitates the effective use of off-the-shelf conic programming solvers, such that real-time performance is attainable. Various robot navigation tasks are investigated to demonstrate the effectiveness of the proposed approach in terms of safety and tracking performance.

cs.RO↗
arXiv2023

Biomedical image analysis competitions: The state of current participation practice

The number of international benchmarking competitions is steadily increasing in various fields of machine learning (ML) research and practice. So far, however, little is known about the common practice as well as bottlenecks faced by the community in tackling the research questions posed. To shed light on the status quo of algorithm development in the specific field of biomedical imaging analysis, we designed an international survey that was issued to all participants of challenges conducted in conjunction with the IEEE ISBI 2021 and MICCAI 2021 conferences (80 competitions in total). The survey covered participants' expertise and working environments, their chosen strategies, as well as algorithm characteristics. A median of 72% challenge participants took part in the survey. According to our results, knowledge exchange was the primary incentive (70%) for participation, while the reception of prize money played only a minor role (16%). While a median of 80 working hours was spent on method development, a large portion of participants stated that they did not have enough time for method development (32%). 25% perceived the infrastructure to be a bottleneck. Overall, 94% of all solutions were deep learning-based. Of these, 84% were based on standard architectures. 43% of the respondents reported that the data samples (e.g., images) were too large to be processed at once. This was most commonly addressed by patch-based training (69%), downsampling (37%), and solving 3D analysis tasks as a series of 2D tasks. K-fold cross-validation on the training set was performed by only 37% of the participants and only 50% of the participants performed ensembling based on multiple identical models (61%) or heterogeneous models (39%). 48% of the respondents applied postprocessing steps.

cs.CV↗
arXiv2023

Real-Time Parallel Trajectory Optimization with Spatiotemporal Safety Constraints for Autonomous Driving in Congested Traffic

Multi-modal behaviors exhibited by surrounding vehicles (SVs) can typically lead to traffic congestion and reduce the travel efficiency of autonomous vehicles (AVs) in dense traffic. This paper proposes a real-time parallel trajectory optimization method for the AV to achieve high travel efficiency in dynamic and congested environments. A spatiotemporal safety module is developed to facilitate the safe interaction between the AV and SVs in the presence of trajectory prediction errors resulting from the multi-modal behaviors of the SVs. By leveraging multiple shooting and constraint transcription, we transform the trajectory optimization problem into a nonlinear programming problem, which allows for the use of optimization solvers and parallel computing techniques to generate multiple feasible trajectories in parallel. Subsequently, these spatiotemporal trajectories are fed into a multi-objective evaluation module considering both safety and efficiency objectives, such that the optimal feasible trajectory corresponding to the optimal target lane can be selected. The proposed framework is validated through simulations in a dense and congested driving scenario with multiple uncertain SVs. The results demonstrate that our method enables the AV to safely navigate through a dense and congested traffic scenario while achieving high travel efficiency and task accuracy in real time.

cs.RO↗
arXiv2023

Existence, uniqueness and ergodicity for McKean-Vlasov SDEs under distribution-dependent Lyapunov conditions

In this paper, we prove the existence and uniqueness of solutions as well as ergodicity for McKean-Vlasov SDEs under Lyapunov conditions, in which the Lyapunov functions are defined on $\mathbb R^d\times \mathcal P_2(\mathbb R^d)$, i.e. the Lyapunov functions depend not only on space variable but also on distribution variable. It is reasonable and natural to consider distribution-dependent Lyapunov functions since the coefficients depends on distribution variable. We apply the martingale representation theorem and a modified Yamada-Watanabe theorem to obtain the existence and uniqueness of solutions. Furthermore, the Krylov-Bogolioubov theorem is used to get ergodicity since it is valid by linearity of the corresponding Fokker-Planck equations on $\mathbb R^d\times \mathcal P_2(\mathbb R^d)$. In particular, if the Lyapunov function depends only on space variable, we obtain exponential ergodicity for semigroup $P_t^*$ under Wasserstein quasi-distance. Finally, we give some examples to illustrate our theoretical results.

math.PR↗
arXiv2023

Curriculum Proximal Policy Optimization with Stage-Decaying Clipping for Self-Driving at Unsignalized Intersections

Unsignalized intersections are typically considered as one of the most representative and challenging scenarios for self-driving vehicles. To tackle autonomous driving problems in such scenarios, this paper proposes a curriculum proximal policy optimization (CPPO) framework with stage-decaying clipping. By adjusting the clipping parameter during different stages of training through proximal policy optimization (PPO), the vehicle can first rapidly search for an approximate optimal policy or its neighborhood with a large parameter, and then converges to the optimal policy with a small one. Particularly, the stage-based curriculum learning technology is incorporated into the proposed framework to improve the generalization performance and further accelerate the training process. Moreover, the reward function is specially designed in view of different curriculum settings. A series of comparative experiments are conducted in intersection-crossing scenarios with bi-lane carriageways to verify the effectiveness of the proposed CPPO method. The results show that the proposed approach demonstrates better adaptiveness to different dynamic and complex environments, as well as faster training speed over baseline methods.

cs.RO↗
arXiv2023

Wireless Network Scheduling with Discrete Propagation Delays: Theorems and Algorithms

This paper focuses on the link scheduling problem in networks where signal delays between nodes are multiples of a time interval. To model such networks, a directed hypergraph is employed, along with an integer matrix that specifies the delays. The link scheduling problem is closely connected to the independent sets of the periodic hypergraph induced by the network model. However, due to the infinite number of vertices, it is impractical to enumerate the independent sets of the periodic hypergraph using generic graph algorithms. To tackle this challenge, a graphical approach is proposed in this paper. The link scheduling rate region is characterized using a finite directed graph called a scheduling graph, which is derived from the network model. A collision-free schedule of the network corresponds to a path in the scheduling graph, and the rate region is determined by the convex hull of the rate vectors associated with the cycles in the scheduling graph. Although existing cycle enumeration algorithms can be employed to calculate the rate region, their computational complexity becomes prohibitively high as the size of the scheduling graph grows exponentially with the number of network links. To address this issue, the dominance property of a special scheduling graph called the step-T scheduling graph is investigated. This property allows the utilization of specific subgraphs of the step-T scheduling graph to characterize the scheduling rate region, achieving a reduction in both the number of cycles and their lengths. For common problems such as calculating the rate region and maximizing a weighted sum of the scheduling rates, algorithms leveraging the dominance property are developed. These algorithms can be more efficient than using generic graph algorithms directly on the scheduling graphs.

cs.IT↗
arXiv2023

Unleashing the Strengths of Unlabeled Data in Pan-cancer Abdominal Organ Quantification: the FLARE22 Challenge

Quantitative organ assessment is an essential step in automated abdominal disease diagnosis and treatment planning. Artificial intelligence (AI) has shown great potential to automatize this process. However, most existing AI algorithms rely on many expert annotations and lack a comprehensive evaluation of accuracy and efficiency in real-world multinational settings. To overcome these limitations, we organized the FLARE 2022 Challenge, the largest abdominal organ analysis challenge to date, to benchmark fast, low-resource, accurate, annotation-efficient, and generalized AI algorithms. We constructed an intercontinental and multinational dataset from more than 50 medical groups, including Computed Tomography (CT) scans with different races, diseases, phases, and manufacturers. We independently validated that a set of AI algorithms achieved a median Dice Similarity Coefficient (DSC) of 90.0\% by using 50 labeled scans and 2000 unlabeled scans, which can significantly reduce annotation requirements. The best-performing algorithms successfully generalized to holdout external validation sets, achieving a median DSC of 89.5\%, 90.9\%, and 88.3\% on North American, European, and Asian cohorts, respectively. They also enabled automatic extraction of key organ biology features, which was labor-intensive with traditional manual measurements. This opens the potential to use unlabeled data to boost performance and alleviate annotation shortages for modern AI models.

eess.IV↗
arXiv2023

Discovery of the shell structure via break radii in the outer halo of the Milky Way

Based on the \textit{Gaia} DR3 RR Lyrae catalog, we use two methods to fit the density profiles with an improved broken power law, and find that there are two break radii coinciding with the two apocenter pile-ups of high-eccentricity Gaia-Sausage-Enceladus (GSE) merger. Also, there is a break caused by the Sagittarius (Sgr) stream. Combining the positions of all breaks, we briefly analyze the metallicity and its dispersion as a function of $r$ as well as its distribution in cylindrical coordinates. For the clean sample, the $z\text{-to-}x$ ellipsoid axial ratio $q$ in $36\,{\rm kpc}\,\textless\,r\,\textless\,96\,{\rm kpc}$ becomes much smaller than that of the inner halo $(r\,\textless\,36\,{\rm kpc})$, while the major axis has a large uncertainty in the region of $36-66\,{\rm kpc}$ and the one in the region of $66-96\,{\rm kpc}$ is obviously different from that dominated by the Hercules-Aquila Cloud (HAC) and the Virgo Overdensity (VOD) in the inner halo, which indicates that there is an over-density structure distributed at low zenithal angles. Finally, we found that the over-density structure in the outer halo ($r\,\textgreater\,50\,{\rm kpc}$) is shell-shaped and relatively metal-rich compared to the outer background halo. We conclude that the shells could be the apocenter pile-ups of the high-eccentricity GSE merger, which is supported by previous numerical simulations.

astro-ph.GA↗
arXiv2023

Information Retrieval Meets Large Language Models: A Strategic Report from Chinese IR Community

The research field of Information Retrieval (IR) has evolved significantly, expanding beyond traditional search to meet diverse user information needs. Recently, Large Language Models (LLMs) have demonstrated exceptional capabilities in text understanding, generation, and knowledge inference, opening up exciting avenues for IR research. LLMs not only facilitate generative retrieval but also offer improved solutions for user understanding, model evaluation, and user-system interactions. More importantly, the synergistic relationship among IR models, LLMs, and humans forms a new technical paradigm that is more powerful for information seeking. IR models provide real-time and relevant information, LLMs contribute internal knowledge, and humans play a central role of demanders and evaluators to the reliability of information services. Nevertheless, significant challenges exist, including computational costs, credibility concerns, domain-specific limitations, and ethical considerations. To thoroughly discuss the transformative impact of LLMs on IR research, the Chinese IR community conducted a strategic workshop in April 2023, yielding valuable insights. This paper provides a summary of the workshop's outcomes, including the rethinking of IR's core values, the mutual enhancement of LLMs and IR, the proposal of a novel IR technical paradigm, and open challenges.

cs.IR↗
arXiv2023

Reinforcement Learning for Robot Navigation with Adaptive Forward Simulation Time (AFST) in a Semi-Markov Model

Deep reinforcement learning (DRL) algorithms have proven effective in robot navigation, especially in unknown environments, by directly mapping perception inputs into robot control commands. However, most existing methods ignore the local minimum problem in navigation and thereby cannot handle complex unknown environments. In this paper, we propose the first DRL-based navigation method modeled by a semi-Markov decision process (SMDP) with continuous action space, named Adaptive Forward Simulation Time (AFST), to overcome this problem. Specifically, we reduce the dimensions of the action space and improve the distributed proximal policy optimization (DPPO) algorithm for the specified SMDP problem by modifying its GAE to better estimate the policy gradient in SMDPs. Experiments in various unknown environments demonstrate the effectiveness of AFST.

cs.RO↗
arXiv2023

A physics-constrained machine learning method for mapping gapless land surface temperature

More accurate, spatio-temporally, and physically consistent LST estimation has been a main interest in Earth system research. Developing physics-driven mechanism models and data-driven machine learning (ML) models are two major paradigms for gapless LST estimation, which have their respective advantages and disadvantages. In this paper, a physics-constrained ML model, which combines the strengths in the mechanism model and ML model, is proposed to generate gapless LST with physical meanings and high accuracy. The hybrid model employs ML as the primary architecture, under which the input variable physical constraints are incorporated to enhance the interpretability and extrapolation ability of the model. Specifically, the light gradient-boosting machine (LGBM) model, which uses only remote sensing data as input, serves as the pure ML model. Physical constraints (PCs) are coupled by further incorporating key Community Land Model (CLM) forcing data (cause) and CLM simulation data (effect) as inputs into the LGBM model. This integration forms the PC-LGBM model, which incorporates surface energy balance (SEB) constraints underlying the data in CLM-LST modeling within a biophysical framework. Compared with a pure physical method and pure ML methods, the PC-LGBM model improves the prediction accuracy and physical interpretability of LST. It also demonstrates a good extrapolation ability for the responses to extreme weather cases, suggesting that the PC-LGBM model enables not only empirical learning from data but also rationally derived from theory. The proposed method represents an innovative way to map accurate and physically interpretable gapless LST, and could provide insights to accelerate knowledge discovery in land surface processes and data mining in geographical parameter estimation.

physics.ao-ph↗
arXiv2023

Graph-Aware Language Model Pre-Training on a Large Graph Corpus Can Help Multiple Graph Applications

Model pre-training on large text corpora has been demonstrated effective for various downstream applications in the NLP domain. In the graph mining domain, a similar analogy can be drawn for pre-training graph models on large graphs in the hope of benefiting downstream graph applications, which has also been explored by several recent studies. However, no existing study has ever investigated the pre-training of text plus graph models on large heterogeneous graphs with abundant textual information (a.k.a. large graph corpora) and then fine-tuning the model on different related downstream applications with different graph schemas. To address this problem, we propose a framework of graph-aware language model pre-training (GALM) on a large graph corpus, which incorporates large language models and graph neural networks, and a variety of fine-tuning methods on downstream applications. We conduct extensive experiments on Amazon's real internal datasets and large public datasets. Comprehensive empirical results and in-depth analysis demonstrate the effectiveness of our proposed methods along with lessons learned.

cs.CL↗
arXiv2023

Astrometric Calibration of the Beijing$-$Arizona Sky Survey

We present the astrometric calibration of the Beijing-Arizona Sky Survey (BASS). The BASS astrometry was tied to the International Celestial Reference Frame via the \emph{Gaia} Data Release 2 reference catalog. For effects that were stable throughout the BASS observations, including differential chromatic refraction and the low charge transfer efficiency of the CCD, we corrected for these effects at the raw image coordinates. Fourth-order polynomial intermediate longitudinal and latitudinal corrections were used to remove optical distortions. The comparison with the \emph{Gaia} catalog shows that the systematic errors, depending on color or magnitude, are less than 2 milliarcseconds (mas). The position systematic error is estimated to be about $-0.01\pm0.7$ mas in the region between 30 and 60 degrees of declination and up to $-0.07 \pm 0.9$ mas in the region north of declination 60 degrees.

astro-ph.IM↗
arXiv2023

Contrastive State Augmentations for Reinforcement Learning-Based Recommender Systems

Learning reinforcement learning (RL)-based recommenders from historical user-item interaction sequences is vital to generate high-reward recommendations and improve long-term cumulative benefits. However, existing RL recommendation methods encounter difficulties (i) to estimate the value functions for states which are not contained in the offline training data, and (ii) to learn effective state representations from user implicit feedback due to the lack of contrastive signals. In this work, we propose contrastive state augmentations (CSA) for the training of RL-based recommender systems. To tackle the first issue, we propose four state augmentation strategies to enlarge the state space of the offline data. The proposed method improves the generalization capability of the recommender by making the RL agent visit the local state regions and ensuring the learned value functions are similar between the original and augmented states. For the second issue, we propose introducing contrastive signals between augmented states and the state randomly sampled from other sessions to improve the state representation learning further. To verify the effectiveness of the proposed CSA, we conduct extensive experiments on two publicly accessible datasets and one dataset collected from a real-life e-commerce platform. We also conduct experiments on a simulated environment as the online evaluation setting. Experimental results demonstrate that CSA can effectively improve recommendation performance.

cs.IR↗