arXiv subjects

Rui Xu

Publications and source records attributed to Rui Xu.

At least 163 records · Page 9Linked to original sources

arXiv2024

RomniStereo: Recurrent Omnidirectional Stereo Matching

Omnidirectional stereo matching (OSM) is an essential and reliable means for $360^{\circ}$ depth sensing. However, following earlier works on conventional stereo matching, prior state-of-the-art (SOTA) methods rely on a 3D encoder-decoder block to regularize the cost volume, causing the whole system complicated and sub-optimal results. Recently, the Recurrent All-pairs Field Transforms (RAFT) based approach employs the recurrent update in 2D and has efficiently improved image-matching tasks, ie, optical flow, and stereo matching. To bridge the gap between OSM and RAFT, we mainly propose an opposite adaptive weighting scheme to seamlessly transform the outputs of spherical sweeping of OSM into the required inputs for the recurrent update, thus creating a recurrent omnidirectional stereo matching (RomniStereo) algorithm. Furthermore, we introduce two techniques, ie, grid embedding and adaptive context feature generation, which also contribute to RomniStereo's performance. Our best model improves the average MAE metric by 40.7\% over the previous SOTA baseline across five datasets. When visualizing the results, our models demonstrate clear advantages on both synthetic and realistic examples. The code is available at \url{https://github.com/HalleyJiang/RomniStereo}.

cs.CV↗

arXiv2024

MAEDiff: Masked Autoencoder-enhanced Diffusion Models for Unsupervised Anomaly Detection in Brain Images

Unsupervised anomaly detection has gained significant attention in the field of medical imaging due to its capability of relieving the costly pixel-level annotation. To achieve this, modern approaches usually utilize generative models to produce healthy references of the diseased images and then identify the abnormalities by comparing the healthy references and the original diseased images. Recently, diffusion models have exhibited promising potential for unsupervised anomaly detection in medical images for their good mode coverage and high sample quality. However, the intrinsic characteristics of the medical images, e.g. the low contrast, and the intricate anatomical structure of the human body make the reconstruction challenging. Besides, the global information of medical images often remain underutilized. To address these two issues, we propose a novel Masked Autoencoder-enhanced Diffusion Model (MAEDiff) for unsupervised anomaly detection in brain images. The MAEDiff involves a hierarchical patch partition. It generates healthy images by overlapping upper-level patches and implements a mechanism based on the masked autoencoders operating on the sub-level patches to enhance the condition on the unnoised regions. Extensive experiments on data of tumors and multiple sclerosis lesions demonstrate the effectiveness of our method.

eess.IV↗

arXiv2023

A Task-driven Network for Mesh Classification and Semantic Part Segmentation

With the rapid development of geometric deep learning techniques, many mesh-based convolutional operators have been proposed to bridge irregular mesh structures and popular backbone networks. In this paper, we show that while convolutions are helpful, a simple architecture based exclusively on multi-layer perceptrons (MLPs) is competent enough to deal with mesh classification and semantic segmentation. Our new network architecture, named Mesh-MLP, takes mesh vertices equipped with the heat kernel signature (HKS) and dihedral angles as the input, replaces the convolution module of a ResNet with Multi-layer Perceptron (MLP), and utilizes layer normalization (LN) to perform the normalization of the layers. The all-MLP architecture operates in an end-to-end fashion and does not include a pooling module. Extensive experimental results on the mesh classification/segmentation tasks validate the effectiveness of the all-MLP architecture.

cs.CV↗

arXiv2023

Probing the vector charge of Sagittarius A* with pulsar timing

Timing a pulsar orbiting around Sagittarius A* (Sgr A*) can provide us with a unique opportunity of testing gravity theories. We investigate the detectability of a vector charge carried by the Sgr A* black hole (BH) in the bumblebee gravity model with simulated future pulsar timing observations. The spacetime of a bumblebee BH introduces characteristic changes to the orbital dynamics of the pulsar and the light propagation of radio signals. Assuming a timing precision of 1 ms, our simulation shows that a 5-yr observation of a pulsar with an orbital period $P_b\sim 0.5\,{\rm yr}$ and an orbital eccentricity $e\sim 0.8$ can probe a vector charge-to-mass ratio as small as $Q/M\sim 10^{-3}$, which is much more stringent than the current constraint from the Event Horizon Telescope (EHT) observations, and comparable to the prospective constraint from extreme mass-ratio inspirals with the Laser Interferometer Space Antenna (LISA).

astro-ph.HE↗

arXiv2023

SCAAT: Improving Neural Network Interpretability via Saliency Constrained Adaptive Adversarial Training

Deep Neural Networks (DNNs) are expected to provide explanation for users to understand their black-box predictions. Saliency map is a common form of explanation illustrating the heatmap of feature attributions, but it suffers from noise in distinguishing important features. In this paper, we propose a model-agnostic learning method called Saliency Constrained Adaptive Adversarial Training (SCAAT) to improve the quality of such DNN interpretability. By constructing adversarial samples under the guidance of saliency map, SCAAT effectively eliminates most noise and makes saliency maps sparser and more faithful without any modification to the model architecture. We apply SCAAT to multiple DNNs and evaluate the quality of the generated saliency maps on various natural and pathological image datasets. Evaluations on different domains and metrics show that SCAAT significantly improves the interpretability of DNNs by providing more faithful saliency maps without sacrificing their predictive power.

cs.CV↗

arXiv2023

Improving Vision-and-Language Reasoning via Spatial Relations Modeling

Visual commonsense reasoning (VCR) is a challenging multi-modal task, which requires high-level cognition and commonsense reasoning ability about the real world. In recent years, large-scale pre-training approaches have been developed and promoted the state-of-the-art performance of VCR. However, the existing approaches almost employ the BERT-like objectives to learn multi-modal representations. These objectives motivated from the text-domain are insufficient for the excavation on the complex scenario of visual modality. Most importantly, the spatial distribution of the visual objects is basically neglected. To address the above issue, we propose to construct the spatial relation graph based on the given visual scenario. Further, we design two pre-training tasks named object position regression (OPR) and spatial relation classification (SRC) to learn to reconstruct the spatial relation graph respectively. Quantitative analysis suggests that the proposed method can guide the representations to maintain more spatial context and facilitate the attention on the essential visual regions for reasoning. We achieve the state-of-the-art results on VCR and two other vision-and-language reasoning tasks VQA, and NLVR.

cs.CV↗

arXiv2023

Non-linear optics at twist interfaces in h-BN/SiC heterostructures

Understanding the emergent electronic structure in twisted atomically thin layers has led to the exciting field of twistronics. However, practical applications of such systems are challenging since the specific angular correlations between the layers must be precisely controlled and the layers have to be single crystalline with uniform atomic ordering. Here, we suggest an alternative, simple and scalable approach where nanocrystalline two-dimensional (2D) film on three-dimensional (3D) substrates yield twisted-interface-dependent properties. Ultrawide-bandgap hexagonal boron nitride (h-BN) thin films are directly grown on high in-plane lattice mismatched wide-bandgap silicon carbide (4H-SiC) substrates to explore the twist-dependent structure-property correlations. Concurrently, nanocrystalline h-BN thin film shows strong non-linear second-harmonic generation and ultra-low cross-plane thermal conductivity at room temperature, which are attributed to the twisted domain edges between van der Waals stacked nanocrystals with random in-plane orientations. First-principles calculations based on time-dependent density functional theory manifest strong even-order optical nonlinearity in twisted h-BN layers. Our work unveils that directly deposited 2D nanocrystalline thin film on 3D substrates could provide easily accessible twist-interfaces, therefore enabling a simple and scalable approach to utilize the 2D-twistronics integrated in 3D material devices for next-generation nanotechnology.

cond-mat.mtrl-sci↗

arXiv2023

Assessing and Enhancing Robustness of Deep Learning Models with Corruption Emulation in Digital Pathology

Deep learning in digital pathology brings intelligence and automation as substantial enhancements to pathological analysis, the gold standard of clinical diagnosis. However, multiple steps from tissue preparation to slide imaging introduce various image corruptions, making it difficult for deep neural network (DNN) models to achieve stable diagnostic results for clinical use. In order to assess and further enhance the robustness of the models, we analyze the physical causes of the full-stack corruptions throughout the pathological life-cycle and propose an Omni-Corruption Emulation (OmniCE) method to reproduce 21 types of corruptions quantified with 5-level severity. We then construct three OmniCE-corrupted benchmark datasets at both patch level and slide level and assess the robustness of popular DNNs in classification and segmentation tasks. Further, we explore to use the OmniCE-corrupted datasets as augmentation data for training and experiments to verify that the generalization ability of the models has been significantly enhanced.

eess.IV↗

arXiv2023

What a Whole Slide Image Can Tell? Subtype-guided Masked Transformer for Pathological Image Captioning

Pathological captioning of Whole Slide Images (WSIs), though is essential in computer-aided pathological diagnosis, has rarely been studied due to the limitations in datasets and model training efficacy. In this paper, we propose a new paradigm Subtype-guided Masked Transformer (SGMT) for pathological captioning based on Transformers, which treats a WSI as a sequence of sparse patches and generates an overall caption sentence from the sequence. An accompanying subtype prediction is introduced into SGMT to guide the training process and enhance the captioning accuracy. We also present an Asymmetric Masked Mechansim approach to tackle the large size constraint of pathological image captioning, where the numbers of sequencing patches in SGMT are sampled differently in the training and inferring phases, respectively. Experiments on the PatchGastricADC22 dataset demonstrate that our approach effectively adapts to the task with a transformer-based model and achieves superior performance than traditional RNN-based methods. Our codes are to be made available for further research and development.

cs.CV↗

arXiv2023

Moment of Inertia for Axisymmetric Neutron Stars in the Standard-Model Extension

We develop a consistent approach to calculate the moment of inertia (MOI) for axisymmetric neutron stars (NSs) in the Lorentz-violating Standard-Model Extension (SME) framework. To our knowledge, this is the first relativistic MOI calculation for axisymmetric NSs in a Lorentz-violating gravity theory other than deformed, rotating NSs in the General Relativity. Under Lorentz violation, there is a specific direction in the spacetime and NSs get stretched or compressed along that direction. When a NS is spinning stationarily along this direction, a conserved angular momentum and the concept of MOI are well defined. In the SME framework, we calculate the partial differential equation governing the rotation and solve it numerically with the finite element method to get the MOI for axisymmetric NSs caused by Lorentz violation. Besides, we study an approximate case where the correction to the MOI is regarded solely from the deformation of the NS and compare it with its counterpart in the Newtonian gravity. Our formalism and the numerical method can be extended to other theories of gravity for static axisymmetric NSs.

gr-qc↗

arXiv2023

Discovering Fatigued Movements for Virtual Character Animation

Virtual character animation and movement synthesis have advanced rapidly during recent years, especially through a combination of extensive motion capture datasets and machine learning. A remaining challenge is interactively simulating characters that fatigue when performing extended motions, which is indispensable for the realism of generated animations. However, capturing such movements is problematic, as performing movements like backflips with fatigued variations up to exhaustion raises capture cost and risk of injury. Surprisingly, little research has been done on faithful fatigue modeling. To address this, we propose a deep reinforcement learning-based approach, which -- for the first time in literature -- generates control policies for full-body physically simulated agents aware of cumulative fatigue. For this, we first leverage Generative Adversarial Imitation Learning (GAIL) to learn an expert policy for the skill; Second, we learn a fatigue policy by limiting the generated constant torque bounds based on endurance time to non-linear, state- and time-dependent limits in the joint-actuation space using a Three-Compartment Controller (3CC) model. Our results demonstrate that agents can adapt to different fatigue and rest rates interactively, and discover realistic recovery strategies without the need for any captured data of fatigued movement.

cs.GR↗

arXiv2023

MiChao-HuaFen 1.0: A Specialized Pre-trained Corpus Dataset for Domain-specific Large Models

With the advancement of deep learning technologies, general-purpose large models such as GPT-4 have demonstrated exceptional capabilities across various domains. Nevertheless, there remains a demand for high-quality, domain-specific outputs in areas like healthcare, law, and finance. This paper first evaluates the existing large models for specialized domains and discusses their limitations. To cater to the specific needs of certain domains, we introduce the ``MiChao-HuaFen 1.0'' pre-trained corpus dataset, tailored for the news and governmental sectors. The dataset, sourced from publicly available internet data from 2022, underwent multiple rounds of cleansing and processing to ensure high quality and reliable origins, with provisions for consistent and stable updates. This dataset not only supports the pre-training of large models for Chinese vertical domains but also aids in propelling deep learning research and applications in related fields.

cs.CL↗

arXiv2023

An Adaptive Federated Relevance Framework for Spatial Temporal Graph Learning

Spatial-temporal data contains rich information and has been widely studied in recent years due to the rapid development of relevant applications in many fields. For instance, medical institutions often use electrodes attached to different parts of a patient to analyse the electorencephal data rich with spatial and temporal features for health assessment and disease diagnosis. Existing research has mainly used deep learning techniques such as convolutional neural network (CNN) or recurrent neural network (RNN) to extract hidden spatial-temporal features. Yet, it is challenging to incorporate both inter-dependencies spatial information and dynamic temporal changes simultaneously. In reality, for a model that leverages these spatial-temporal features to fulfil complex prediction tasks, it often requires a colossal amount of training data in order to obtain satisfactory model performance. Considering the above-mentioned challenges, we propose an adaptive federated relevance framework, namely FedRel, for spatial-temporal graph learning in this paper. After transforming the raw spatial-temporal data into high quality features, the core Dynamic Inter-Intra Graph (DIIG) module in the framework is able to use these features to generate the spatial-temporal graphs capable of capturing the hidden topological and long-term temporal correlation information in these graphs. To improve the model generalization ability and performance while preserving the local data privacy, we also design a relevance-driven federated learning module in our framework to leverage diverse data distributions from different participants with attentive aggregations of their models.

cs.LG↗

arXiv2023

Neural-Singular-Hessian: Implicit Neural Representation of Unoriented Point Clouds by Enforcing Singular Hessian

Neural implicit representation is a promising approach for reconstructing surfaces from point clouds. Existing methods combine various regularization terms, such as the Eikonal and Laplacian energy terms, to enforce the learned neural function to possess the properties of a Signed Distance Function (SDF). However, inferring the actual topology and geometry of the underlying surface from poor-quality unoriented point clouds remains challenging. In accordance with Differential Geometry, the Hessian of the SDF is singular for points within the differential thin-shell space surrounding the surface. Our approach enforces the Hessian of the neural implicit function to have a zero determinant for points near the surface. This technique aligns the gradients for a near-surface point and its on-surface projection point, producing a rough but faithful shape within just a few iterations. By annealing the weight of the singular-Hessian term, our approach ultimately produces a high-fidelity reconstruction result. Extensive experimental results demonstrate that our approach effectively suppresses ghost geometry and recovers details from unoriented point clouds with better expressiveness than existing fitting-based methods.

cs.CV↗

arXiv2023

Rethinking the Localization in Weakly Supervised Object Localization

Weakly supervised object localization (WSOL) is one of the most popular and challenging tasks in computer vision. This task is to localize the objects in the images given only the image-level supervision. Recently, dividing WSOL into two parts (class-agnostic object localization and object classification) has become the state-of-the-art pipeline for this task. However, existing solutions under this pipeline usually suffer from the following drawbacks: 1) they are not flexible since they can only localize one object for each image due to the adopted single-class regression (SCR) for localization; 2) the generated pseudo bounding boxes may be noisy, but the negative impact of such noise is not well addressed. To remedy these drawbacks, we first propose to replace SCR with a binary-class detector (BCD) for localizing multiple objects, where the detector is trained by discriminating the foreground and background. Then we design a weighted entropy (WE) loss using the unlabeled data to reduce the negative impact of noisy bounding boxes. Extensive experiments on the popular CUB-200-2011 and ImageNet-1K datasets demonstrate the effectiveness of our method.

cs.CV↗

arXiv2023

Relay-Enabled Backscatter Communications: Linear Mapping and Resource Allocation

Relay-enabled backscatter communication (BC) is an intriguing paradigm to alleviate energy shortage and improve throughput of Internet-of-Things (IoT) devices. Most of the existing works focus on the resource allocation that considered the unequal and continuous time allocation for both source-relay and relay-destination links. However, the continuous time allocation may be infeasible since in practice, the time allocation shall be carried out in integral multiple of the subframe duration unit. In this article, we study a discrete time scheme from the perspective of frame structure, where one transmission block is divided into two phases and the linear mapping is employed as a re-encoding method to determine the number of subframes for both phases and the power allocation for each subframe in a relay-enabled BC system. Based on this, we derive an accurate system-throughput expression and formulate a mixed-integral non-convex optimization problem to maximize the system throughput by jointly optimizing the power reflection coefficient (PRC) of the IoT node, the power allocation of the hybrid access point (HAP) and the linear mapping matrix, and solve it via a three-step approach. Accordingly, we propose a low complexity iterative algorithm to obtain the throughput maximization-based resource allocation solution. Numerical results analyze the performance of our proposed algorithm, verify the superiority of our proposed scheme, and evaluate the impacts of network parameters on the system throughput.

cs.IT↗

arXiv2023

Extended thermodynamics of the bumblebee black holes

As a vector-tensor theory including nonminimal coupling between the Ricci tensor and a vector field, the bumblebee gravity is a potential theory to test Lorentz symmetry violation. Recently, a new class of numerical spherical black holes in the bumblebee theory was constructed. In this paper, we investigate the associated local thermodynamic properties. By introducing a pair of conjugated thermodynamic quantities $X$ and $Y$, which can be interpreted as an extension of electric potential and charge of the Reissner Nordström black holes, we numerically construct a new first law of thermodynamics for bumblebee black holes. We then study the constant-$Y$ processes in the entropy-charge parameter space. For the constant-$Y$ processes, we also calculate the heat capacity to study the local thermodynamic stability of the bumblebee black holes. For a negative nonminimal coupling coefficient $ξ$, we find both divergent and smooth phase transitions. For a positive but small $ξ$, only a divergent phase transition is found. It turns out that there is a critical value $0.4κ<ξ_c < 0.5κ$ such that when $ξ_c < ξ<2κ$, even the divergent phase transition disappears and the bumblebee black holes thus become locally thermodynamically unstable regardless of the bumblebee charge. As for $ξ>2κ$, the smooth phase transition arises again but there no longer exists any discontinuous phase transition for the bumblebee black holes.

gr-qc↗

arXiv2023

Determination of geopotential difference by hydrogen masers based on precise point positioning time-frequency transfer

According to the general relativity theory, the geopotential difference can be determined by gravity frequency shift between two clocks. Here we report on the experiments to determine the geopotential difference between two remote sites by hydrogen masers based on precise point positioning time-frequency transfer technique. The experiments include the remote clock comparison and the local clock comparison using two CH1-95 active hydrogen masers linked with global navigation satellite system time-frequency receivers. The frequency difference between two hydrogen masers at two sites is derived from the time difference series resolved by the above-mentioned technique. Considering the local clock comparison as calibration, the determined geopotential difference by our experiments is 12,142.3 (112.4) m^2/s^2, quite close to the value 12,153.3 (2.3) m^2/s^2 computed by the EIGEN-6C4 model. Results show that the proposed approach here for determining geopotential difference is feasible, operable, and promising for applications in various fields.

physics.geo-ph↗