arXiv subjects

Hui Zhang

Publications and source records attributed to Hui Zhang.

At least 235 records · Page 13Linked to original sources

arXiv2024

Is Your HD Map Constructor Reliable under Sensor Corruptions?

Driving systems often rely on high-definition (HD) maps for precise environmental information, which is crucial for planning and navigation. While current HD map constructors perform well under ideal conditions, their resilience to real-world challenges, \eg, adverse weather and sensor failures, is not well understood, raising safety concerns. This work introduces MapBench, the first comprehensive benchmark designed to evaluate the robustness of HD map construction methods against various sensor corruptions. Our benchmark encompasses a total of 29 types of corruptions that occur from cameras and LiDAR sensors. Extensive evaluations across 31 HD map constructors reveal significant performance degradation of existing methods under adverse weather conditions and sensor failures, underscoring critical safety concerns. We identify effective strategies for enhancing robustness, including innovative approaches that leverage multi-modal fusion, advanced data augmentation, and architectural techniques. These insights provide a pathway for developing more reliable HD map construction methods, which are essential for the advancement of autonomous driving technology. The benchmark toolkit and affiliated code and model checkpoints have been made publicly accessible.

cs.RO↗

arXiv2024

Edge-Cloud Collaborative Satellite Image Analysis for Efficient Man-Made Structure Recognition

The increasing availability of high-resolution satellite imagery has created immense opportunities for various applications. However, processing and analyzing such vast amounts of data in a timely and accurate manner poses significant challenges. The paper presents a new satellite image processing architecture combining edge and cloud computing to better identify man-made structures against natural landscapes. By employing lightweight models at the edge, the system initially identifies potential man-made structures from satellite imagery. These identified images are then transmitted to the cloud, where a more complex model refines the classification, determining specific types of structures. The primary focus is on the trade-off between latency and accuracy, as efficient models often sacrifice accuracy. We compare this hybrid edge-cloud approach against traditional "bent-pipe" method in virtual environment experiments as well as introduce a practical model and compare its performance with existing lightweight models for edge deployment, focusing on accuracy and latency. The results demonstrate that the edge-cloud collaborative model not only reduces overall latency due to minimized data transmission but also maintains high accuracy, offering substantial improvements over traditional approaches under this scenario.

cs.CV↗

arXiv2024

Highly coherent grain boundaries induced by local pseudo-mirror symmetry in $β$-Ga2O3

Grain boundaries have extensive influence on the performance of crystal materials. However, the atomic-scale structure and its relation with local and crystallographic symmetries remain elusive in low-symmetry crystals. Herein, we find that the local pseudo-mirror-symmetric atomic layer is the common physical origin of a series of highly coherent grain boundaries in the low-symmetry $β$-Ga2O3 crystal. These include the (100) twin boundary and an emerging series of $(h-1'0'2)/(h+1'0'\bar{2})$ coherent asymmetric grain boundaries (CAGBs). Owing to the local pseudo-mirror symmetry and the special geometric relation of the $β$-Ga2O3 conventional cell, these CAGBs place 80% of the boundary atoms in pseudo-coincident sites, exhibiting high coherence under the coincident-site lattice model. With a combination of density functional theory calculations, Czochralski growth experiment, and atomic-scale characterizations, the structure and stability of the $(002)/(20\bar{2})$-A CAGB are confirmed, with a boundary energy density as low as 0.36 J/m2. This CAGB is responsible for the spontaneous formation of a twinned defect facet at the surface steps during the epitaxy growth of $β$-Ga2O3, warranting a substrate orientation selection rule for $β$-Ga2O3. Through this study, we provide insights into the grain boundary physics in the low-symmetry $β$-Ga2O3 crystal while emphasizing the importance of the local pseudo-symmetries in the low-symmetry crystals.

cond-mat.mtrl-sci↗

arXiv2024

ToolEENet: Tool Affordance 6D Pose Estimation

The exploration of robotic dexterous hands utilizing tools has recently attracted considerable attention. A significant challenge in this field is the precise awareness of a tool's pose when grasped, as occlusion by the hand often degrades the quality of the estimation. Additionally, the tool's overall pose often fails to accurately represent the contact interaction, thereby limiting the effectiveness of vision-guided, contact-dependent activities. To overcome this limitation, we present the innovative TOOLEE dataset, which, to the best of our knowledge, is the first to feature affordance segmentation of a tool's end-effector (EE) along with its defined 6D pose based on its usage. Furthermore, we propose the ToolEENet framework for accurate 6D pose estimation of the tool's EE. This framework begins by segmenting the tool's EE from raw RGBD data, then uses a diffusion model-based pose estimator for 6D pose estimation at a category-specific level. Addressing the issue of symmetry in pose estimation, we introduce a symmetry-aware pose representation that enhances the consistency of pose estimation. Our approach excels in this field, demonstrating high levels of precision and generalization. Furthermore, it shows great promise for application in contact-based manipulation scenarios. All data and codes are available on the project website: https://tooleenet-iros2024.github.io/

cs.RO↗

arXiv2024

Segmentation-guided Layer-wise Image Vectorization with Gradient Fills

The widespread use of vector graphics creates a significant demand for vectorization methods. While recent learning-based techniques have shown their capability to create vector images of clear topology, filling these primitives with gradients remains a challenge. In this paper, we propose a segmentation-guided vectorization framework to convert raster images into concise vector graphics with radial gradient fills. With the guidance of an embedded gradient-aware segmentation subroutine, our approach progressively appends gradient-filled Bézier paths to the output, where primitive parameters are initiated with our newly designed initialization technique and are optimized to minimize our novel loss function. We build our method on a differentiable renderer with traditional segmentation algorithms to develop it as a model-free tool for raster-to-vector conversion. It is tested on various inputs to demonstrate its feasibility, independent of datasets, to synthesize vector graphics with improved visual quality and layer-wise topology compared to prior work.

cs.CV↗

arXiv2024

La-SoftMoE CLIP for Unified Physical-Digital Face Attack Detection

Facial recognition systems are susceptible to both physical and digital attacks, posing significant security risks. Traditional approaches often treat these two attack types separately due to their distinct characteristics. Thus, when being combined attacked, almost all methods could not deal. Some studies attempt to combine the sparse data from both types of attacks into a single dataset and try to find a common feature space, which is often impractical due to the space is difficult to be found or even non-existent. To overcome these challenges, we propose a novel approach that uses the sparse model to handle sparse data, utilizing different parameter groups to process distinct regions of the sparse feature space. Specifically, we employ the Mixture of Experts (MoE) framework in our model, expert parameters are matched to tokens with varying weights during training and adaptively activated during testing. However, the traditional MoE struggles with the complex and irregular classification boundaries of this problem. Thus, we introduce a flexible self-adapting weighting mechanism, enabling the model to better fit and adapt. In this paper, we proposed La-SoftMoE CLIP, which allows for more flexible adaptation to the Unified Attack Detection (UAD) task, significantly enhancing the model's capability to handle diversity attacks. Experiment results demonstrate that our proposed method has SOTA performance.

cs.CV↗

arXiv2024

Compartment-specific estimation of T2 and T2* with diffusion-PEPTIDE MRI

We present a microstructure imaging technique for estimating compartment-specific T2 and T2* simultaneously in the human brain. Microstructure imaging with diffusion MRI (dMRI) has enabled the modelling of intra-neurite and extra-neurite diffusion signals separately allowing for the estimation of compartment-specific tissue properties. These compartment-specific properties have been widely used in clinical studies. However, conventional dMRI cannot disentangle differences in relaxations between tissue compartments, causing biased estimates of diffusion measures which also change with TE. To solve the problem, combined relaxometry-diffusion imaging methods have been developed in recent years, providing compartmental T2-diffusion or T2*-diffusion imaging respectively, but not T2 and T2* together. As they provide complementary information, a technique that can estimate both jointly with diffusion is appealing to neuroimaging studies. The aim of this work is to develop a method to map compartmental T2-T2*-diffusion simultaneously. Using an advanced MRI acquisition called diffusion-PEPTIDE, a novel microstructure model is proposed and a multi-step fitting method is developed to estimate parameters of interest. We demonstrate for the first time that compartmental T2, T2* can be estimated simultaneously from in vivo data. we further show the accuracy and precision of parameter estimation with simulation.

physics.med-ph↗

arXiv2024

SCREENER: A general framework for task-specific experiment design in quantitative MRI

Quantitative magnetic resonance imaging (qMRI) is increasingly investigated for use in a variety of clinical tasks from diagnosis, through staging, to treatment monitoring. However, experiment design in qMRI, the identification of the optimal acquisition protocols, has been focused on obtaining the most precise parameter estimations, with no regard for the specific requirements of downstream tasks. Here we propose SCREENER: A general framework for task-specific experiment design in quantitative MRI. SCREENER incorporates a task-specific objective and seeks the optimal protocol with a deep-reinforcement-learning (DRL) based optimization strategy. To illustrate this framework, we employ a task of classifying the inflammation status of bone marrow using diffusion MRI data with intravoxel incoherent motion (IVIM) modelling. Results demonstrate SCREENER outperforms previous ad hoc and optimized protocols under clinical signal-to-noise ratio (SNR) conditions, achieving significant improvement, both in binary classification tasks, e.g. from 67% to 89%, and in a multi-class classification task, from 46% to 59%. Additionally, we show this improvement is robust to the SNR. Lastly, we demonstrate the advantage of DRL-based optimization strategy, enabling zero-shot discovery of near-optimal protocols for a range of SNRs not used in training. In conclusion, SCREENER has the potential to enable wider uptake of qMRI in the clinic.

cs.CV↗

arXiv2024

Dynamic Identity-Guided Attention Network for Visible-Infrared Person Re-identification

Visible-infrared person re-identification (VI-ReID) aims to match people with the same identity between visible and infrared modalities. VI-ReID is a challenging task due to the large differences in individual appearance under different modalities. Existing methods generally try to bridge the cross-modal differences at image or feature level, which lacks exploring the discriminative embeddings. Effectively minimizing these cross-modal discrepancies relies on obtaining representations that are guided by identity and consistent across modalities, while also filtering out representations that are irrelevant to identity. To address these challenges, we introduce a dynamic identity-guided attention network (DIAN) to mine identity-guided and modality-consistent embeddings, facilitating effective bridging the gap between different modalities. Specifically, in DIAN, to pursue a semantically richer representation, we first use orthogonal projection to fuse the features from two connected coarse and fine layers. Furthermore, we first use dynamic convolution kernels to mine identity-guided and modality-consistent representations. More notably, a cross embedding balancing loss is introduced to effectively bridge cross-modal discrepancies by above embeddings. Experimental results on SYSU-MM01 and RegDB datasets show that DIAN achieves state-of-the-art performance. Specifically, for indoor search on SYSU-MM01, our method achieves 86.28% rank-1 accuracy and 87.41% mAP, respectively. Our code will be available soon.

cs.CV↗

arXiv2024

WiFaKey: Generating Cryptographic Keys from Face in the Wild

Deriving a unique cryptographic key from biometric measurements is a challenging task due to the existing noise gap between the biometric measurements and error correction coding. Additionally, privacy and security concerns arise as biometric measurements are inherently linked to the user. Biocryptosystems represent a key branch of solutions aimed at addressing these issues. However, many existing bio-cryptosystems rely on handcrafted feature extractors and error correction codes (ECC), often leading to performance degradation. To address these challenges and improve the reliability of biometric measurements, we propose a novel biometric cryptosystem named WiFaKey, for generating cryptographic keys from face in unconstrained settings. Speciffcally, WiFaKey ffrst introduces an adaptive random masking-driven feature transformation pipeline, AdaMTrans. AdaMTrans effectively quantizes and binarizes realvalued features and incorporates an adaptive random masking scheme to align the bit error rate with error correction requirements, thereby mitigating the noise gap. Besides, WiFaKey incorporates a supervised learning-based neural decoding scheme called Neural-MS decoder, which delivers a more robust error correction performance with less iteration than non-learning decoders, thereby alleviating the performance degradation. We evaluated WiFaKey using widely adopted face feature extractors on six large unconstrained and two constrained datasets. On the LFW dataset, WiFaKey achieved an average Genuine Match Rate of 85.45% and 85.20% at a 0% False Match Rate for MagFace and AdaFace features, respectively. Our comprehensive comparative analysis shows a signiffcant performance improvement of WiFaKey. The source code of our work is available at github.com/xingbod/WiFaKey.

cs.CR↗

arXiv2024

Unraveling the multistage phase transformations in monolayer Mo-Te compounds

Monolayer MoTe2 exhibits a variety of derivative structural phases and associated novel electronic properties that enable a wealth of potential applications in future electronic and optoelectronic devices. However, a comprehensive study focusing on the complexities of the controllable phase evolution in this atomically thin film has yet to be performed. This work aims to address this issue by systematically investigating molecular beam epitaxial growth of monolayer Mo-Te compounds on bilayer graphene substrates. By utilizing scanning tunnelling microscopy, we explored a series of thermally driven structural phase evolutions including distinct T'-MoTe2, H-MoTe2, Mo6Te6 nanowires, and multistoichiometric MoTe2-x. Furthermore, we carefully investigated the critical effects of the growth parameters-annealing temperature and time and tellurium concentration-on the controllable and reversible phase transformation within monolayer MoTe2-x. The findings have significant implications for understanding the thin film synthesis and phase transformation engineering inherent to two-dimensional crystals, which can foster further development of high-performance devices.

cond-mat.mtrl-sci↗

arXiv2024

Team Samsung-RAL: Technical Report for 2024 RoboDrive Challenge-Robust Map Segmentation Track

In this report, we describe the technical details of our submission to the 2024 RoboDrive Challenge Robust Map Segmentation Track. The Robust Map Segmentation track focuses on the segmentation of complex driving scene elements in BEV maps under varied driving conditions. Semantic map segmentation provides abundant and precise static environmental information crucial for autonomous driving systems' planning and navigation. While current methods excel in ideal circumstances, e.g., clear daytime conditions and fully functional sensors, their resilience to real-world challenges like adverse weather and sensor failures remains unclear, raising concerns about system safety. In this paper, we explored several methods to improve the robustness of the map segmentation task. The details are as follows: 1) Robustness analysis of utilizing temporal information; 2) Robustness analysis of utilizing different backbones; and 3) Data Augmentation to boost corruption robustness. Based on the evaluation results, we draw several important findings including 1) The temporal fusion module is effective in improving the robustness of the map segmentation model; 2) A strong backbone is effective for improving the corruption robustness; and 3) Some data augmentation methods are effective in improving the robustness of map segmentation models. These novel findings allowed us to achieve promising results in the 2024 RoboDrive Challenge-Robust Map Segmentation Track.

cs.CV↗

arXiv2024

MapDistill: Boosting Efficient Camera-based HD Map Construction via Camera-LiDAR Fusion Model Distillation

Online high-definition (HD) map construction is an important and challenging task in autonomous driving. Recently, there has been a growing interest in cost-effective multi-view camera-based methods without relying on other sensors like LiDAR. However, these methods suffer from a lack of explicit depth information, necessitating the use of large models to achieve satisfactory performance. To address this, we employ the Knowledge Distillation (KD) idea for efficient HD map construction for the first time and introduce a novel KD-based approach called MapDistill to transfer knowledge from a high-performance camera-LiDAR fusion model to a lightweight camera-only model. Specifically, we adopt the teacher-student architecture, i.e., a camera-LiDAR fusion model as the teacher and a lightweight camera model as the student, and devise a dual BEV transform module to facilitate cross-modal knowledge distillation while maintaining cost-effective camera-only deployment. Additionally, we present a comprehensive distillation scheme encompassing cross-modal relation distillation, dual-level feature distillation, and map head distillation. This approach alleviates knowledge transfer challenges between modalities, enabling the student model to learn improved feature representations for HD map construction. Experimental results on the challenging nuScenes dataset demonstrate the effectiveness of MapDistill, surpassing existing competitors by over 7.7 mAP or 4.5X speedup.

cs.CV↗

arXiv2024

GraspXL: Generating Grasping Motions for Diverse Objects at Scale

Human hands possess the dexterity to interact with diverse objects such as grasping specific parts of the objects and/or approaching them from desired directions. More importantly, humans can grasp objects of any shape without object-specific skills. Recent works synthesize grasping motions following single objectives such as a desired approach heading direction or a grasping area. Moreover, they usually rely on expensive 3D hand-object data during training and inference, which limits their capability to synthesize grasping motions for unseen objects at scale. In this paper, we unify the generation of hand-object grasping motions across multiple motion objectives, diverse object shapes and dexterous hand morphologies in a policy learning framework GraspXL. The objectives are composed of the graspable area, heading direction during approach, wrist rotation, and hand position. Without requiring any 3D hand-object interaction data, our policy trained with 58 objects can robustly synthesize diverse grasping motions for more than 500k unseen objects with a success rate of 82.2%. At the same time, the policy adheres to objectives, which enables the generation of diverse grasps per object. Moreover, we show that our framework can be deployed to different dexterous hands and work with reconstructed or generated objects. We quantitatively and qualitatively evaluate our method to show the efficacy of our approach. Our model, code, and the large-scale generated motions are available at https://eth-ait.github.io/graspxl/.

cs.RO↗

arXiv2024

Region Attention Transformer for Medical Image Restoration

Transformer-based methods have demonstrated impressive results in medical image restoration, attributed to the multi-head self-attention (MSA) mechanism in the spatial dimension. However, the majority of existing Transformers conduct attention within fixed and coarsely partitioned regions (\text{e.g.} the entire image or fixed patches), resulting in interference from irrelevant regions and fragmentation of continuous image content. To overcome these challenges, we introduce a novel Region Attention Transformer (RAT) that utilizes a region-based multi-head self-attention mechanism (R-MSA). The R-MSA dynamically partitions the input image into non-overlapping semantic regions using the robust Segment Anything Model (SAM) and then performs self-attention within these regions. This region partitioning is more flexible and interpretable, ensuring that only pixels from similar semantic regions complement each other, thereby eliminating interference from irrelevant regions. Moreover, we introduce a focal region loss to guide our model to adaptively focus on recovering high-difficulty regions. Extensive experiments demonstrate the effectiveness of RAT in various medical image restoration tasks, including PET image synthesis, CT image denoising, and pathological image super-resolution. Code is available at \href{https://github.com/Yaziwel/Region-Attention-Transformer-for-Medical-Image-Restoration.git}{https://github.com/RAT}.

eess.IV↗

arXiv2024

Non-Hermitian Lattice Fermions in 2D GNY Model

We work the lattice fermions and non-Hermitian formulation in the 2D GNY model and demonstrate the numerical implementation for two flavors by the Hybrid Monte Carlo. Our approach has a notable advantage in dealing with chiral symmetry on a lattice by avoiding the Nielsen-Ninomiya theorem, due to the non-symmetrized finite-difference operator. We restore the hypercubic symmetry by averaging over all possible orientations with the proper continuum limit. Our study is the first simulation for the interacting fermion formulated in a non-hermitian way. We compare the numerical solution with the one-loop resummation. The resummation results matches with the numerical solution in $\langleϕ\rangle$, $\langleϕ^2\rangle$, $\langle\mathrm{Tr}(\barψ_1ψ_1+\barψ_2ψ_2)/2\rangle$, and $\langle\mathrm{Tr}(\barψ_1ψ_1+\barψ_2ψ_2)ϕ/2\rangle$. We also used the one-loop resummation to provide the RG flow and asymptotic safety in the 2D GNY model.

hep-th↗

arXiv2024

Not all explicit cues help communicate: Pedestrians' perceptions, fixations, and decisions toward automated vehicles with varied appearance

Given pedestrians' vulnerability in road traffic, it remains unclear how novel AV appearances will impact pedestrians crossing behaviour. To address this gap, this study pioneers an investigation into the influence of AVs' exterior design, correlated with their kinematics, on pedestrians' road-crossing perception and decision-making. A video-based eye-tracking experimental study was conducted with 61 participants who responded to video stimuli depicting a manipulated vehicle approaching a predefined road-crossing location on an unsignalized, two-way road. The vehicle's kinematic pattern was manipulated into yielding and non-yielding, and its external appearances were varied across five types: with a human driver (as a conventional vehicle), with no driver (as an AV), with text-based identity indications, with roof radar sensors, with dynamic eHMIs adjusted to vehicle kinematics. Participants' perceived clarity, crossing initiation distance (CID), crossing decision time (CDT), and gaze behaviour, during interactions were recorded and reported. The results indicated that AVs' kinematic profiles play a dominant role in pedestrians' road-crossing decisions, supported by their subjective evaluations, CID, CDT, and gaze patterns during interactions. Moreover, the use of clear eHMI, such as dynamic pedestrian icons, reduced pedestrians' visual load, enhanced their perceptual clarity, expedited road-crossing decisions, and thereby improved overall crossing efficiency. However, it was found that both textual identity indications and roof radar sensors have no significant effect on pedestrians' decisions but negatively impact pedestrians' visual attention, as evidenced by heightened fixation counts and prolonged fixation durations, particularly under yielding conditions. Excessive visual and cognitive resource occupation suggests that not all explicit cues facilitate human-vehicle communication.

cs.HC↗

arXiv2024

POLAFFINI: Efficient feature-based polyaffine initialization for improved non-linear image registration

This paper presents an efficient feature-based approach to initialize non-linear image registration. Today, nonlinear image registration is dominated by methods relying on intensity-based similarity measures. A good estimate of the initial transformation is essential, both for traditional iterative algorithms and for recent one-shot deep learning (DL)-based alternatives. The established approach to estimate this starting point is to perform affine registration, but this may be insufficient due to its parsimonious, global, and non-bending nature. We propose an improved initialization method that takes advantage of recent advances in DL-based segmentation techniques able to instantly estimate fine-grained regional delineations with state-of-the-art accuracies. Those segmentations are used to produce local, anatomically grounded, feature-based affine matchings using iteration-free closed-form expressions. Estimated local affine transformations are then fused, with the log-Euclidean polyaffine framework, into an overall dense diffeomorphic transformation. We show that, compared to its affine counterpart, the proposed initialization leads to significantly better alignment for both traditional and DL-based non-linear registration algorithms. The proposed approach is also more robust and significantly faster than commonly used affine registration algorithms such as FSL FLIRT.

cs.CV↗