arXiv subjects

Lei Wang

Publications and source records attributed to Lei Wang.

At least 595 records · Page 33Linked to original sources

arXiv2024

Taylor Videos for Action Recognition

Effectively extracting motions from video is a critical and long-standing problem for action recognition. This problem is very challenging because motions (i) do not have an explicit form, (ii) have various concepts such as displacement, velocity, and acceleration, and (iii) often contain noise caused by unstable pixels. Addressing these challenges, we propose the Taylor video, a new video format that highlights the dominate motions (e.g., a waving hand) in each of its frames named the Taylor frame. Taylor video is named after Taylor series, which approximates a function at a given point using important terms. In the scenario of videos, we define an implicit motion-extraction function which aims to extract motions from video temporal block. In this block, using the frames, the difference frames, and higher-order difference frames, we perform Taylor expansion to approximate this function at the starting frame. We show the summation of the higher-order terms in the Taylor series gives us dominant motion patterns, where static objects, small and unstable motions are removed. Experimentally we show that Taylor videos are effective inputs to popular architectures including 2D CNNs, 3D CNNs, and transformers. When used individually, Taylor videos yield competitive action recognition accuracy compared to RGB videos and optical flow. When fused with RGB or optical flow videos, further accuracy improvement is achieved. Additionally, we apply Taylor video computation to human skeleton sequences, resulting in Taylor skeleton sequences that outperform the use of original skeletons for skeleton-based action recognition.

cs.CV↗

arXiv2024

Ditto: Quantization-aware Secure Inference of Transformers upon MPC

Due to the rising privacy concerns on sensitive client data and trained models like Transformers, secure multi-party computation (MPC) techniques are employed to enable secure inference despite attendant overhead. Existing works attempt to reduce the overhead using more MPC-friendly non-linear function approximations. However, the integration of quantization widely used in plaintext inference into the MPC domain remains unclear. To bridge this gap, we propose the framework named Ditto to enable more efficient quantization-aware secure Transformer inference. Concretely, we first incorporate an MPC-friendly quantization into Transformer inference and employ a quantization-aware distillation procedure to maintain the model utility. Then, we propose novel MPC primitives to support the type conversions that are essential in quantization and implement the quantization-aware MPC execution of secure quantized inference. This approach significantly decreases both computation and communication overhead, leading to improvements in overall efficiency. We conduct extensive experiments on Bert and GPT2 models to evaluate the performance of Ditto. The results demonstrate that Ditto is about $3.14\sim 4.40\times$ faster than MPCFormer (ICLR 2023) and $1.44\sim 2.35\times$ faster than the state-of-the-art work PUMA with negligible utility degradation.

cs.CR↗

arXiv2024

In-context Autoencoder for Context Compression in a Large Language Model

We propose the In-context Autoencoder (ICAE), leveraging the power of a large language model (LLM) to compress a long context into short compact memory slots that can be directly conditioned on by the LLM for various purposes. ICAE is first pretrained using both autoencoding and language modeling objectives on massive text data, enabling it to generate memory slots that accurately and comprehensively represent the original context. Then, it is fine-tuned on instruction data for producing desirable responses to various prompts. Experiments demonstrate that our lightweight ICAE, introducing about 1% additional parameters, effectively achieves $4\times$ context compression based on Llama, offering advantages in both improved latency and GPU memory cost during inference, and showing an interesting insight in memorization as well as potential for scalability. These promising results imply a novel perspective on the connection between working memory in cognitive science and representation learning in LLMs, revealing ICAE's significant implications in addressing the long context problem and suggesting further research in LLM context management. Our data, code and models are available at https://github.com/getao/icae.

cs.CL↗

arXiv2024

Differentially Private Average Consensus with Improved Accuracy-Privacy Trade-off

This paper studies the average consensus problem with differential privacy of initial states, for which it is widely recognized that there is a trade-off between the mean-square computation accuracy and privacy level. Considering the trade-off gap between the average consensus algorithm and the centralized averaging approach with differential privacy, we propose a distributed shuffling mechanism based on the Paillier cryptosystem to generate correlated zero-sum randomness. By randomizing each local privacy-sensitive initial state with an i.i.d. Gaussian noise and the output of the mechanism using Gaussian noises, it is shown that the resulting average consensus algorithm can eliminate the gap in the sense that the accuracy-privacy trade-off of the centralized averaging approach with differential privacy can be almost recovered by appropriately designing the variances of the added noises. We also extend such a design framework with Gaussian noises to the one using Laplace noises, and show that the improved privacy-accuracy trade-off is preserved.

eess.SY↗

arXiv2024

A New Self-Alignment Method without Solving Wahba Problem for SINS in Autonomous Vehicles

Initial alignment is one of the key technologies in strapdown inertial navigation system (SINS) to provide initial state information for vehicle attitude and navigation. For some situations, such as the attitude heading reference system, the position is not necessarily required or even available, then the self-alignment that does not rely on any external aid becomes very necessary. This study presents a new self-alignment method under swaying conditions, which can determine the latitude and attitude simultaneously by utilizing all observation vectors without solving the Wahba problem, and it is different from the existing methods. By constructing the dyadic tensor of each observation and reference vector itself, all equations related to observation and reference vectors are accumulated into one equation, where the latitude variable is extracted and solved according to the same eigenvalues of similar matrices on both sides of the equation, meanwhile the attitude is obtained by eigenvalue decomposition. Simulation and experiment tests verify the effectiveness of the proposed methods, and the alignment result is better than TRIAD in convergence speed and stability and comparable with OBA method in alignment accuracy with or without latitude. It is useful for guiding the design of initial alignment in autonomous vehicle applications.

cs.RO↗

arXiv2024

HiH: A Multi-modal Hierarchy in Hierarchy Network for Unconstrained Gait Recognition

Gait recognition has achieved promising advances in controlled settings, yet it significantly struggles in unconstrained environments due to challenges such as view changes, occlusions, and varying walking speeds. Additionally, efforts to fuse multiple modalities often face limited improvements because of cross-modality incompatibility, particularly in outdoor scenarios. To address these issues, we present a multi-modal Hierarchy in Hierarchy network (HiH) that integrates silhouette and pose sequences for robust gait recognition. HiH features a main branch that utilizes Hierarchical Gait Decomposer (HGD) modules for depth-wise and intra-module hierarchical examination of general gait patterns from silhouette data. This approach captures motion hierarchies from overall body dynamics to detailed limb movements, facilitating the representation of gait attributes across multiple spatial resolutions. Complementing this, an auxiliary branch, based on 2D joint sequences, enriches the spatial and temporal aspects of gait analysis. It employs a Deformable Spatial Enhancement (DSE) module for pose-guided spatial attention and a Deformable Temporal Alignment (DTA) module for aligning motion dynamics through learned temporal offsets. Extensive evaluations across diverse indoor and outdoor datasets demonstrate HiH's state-of-the-art performance, affirming a well-balanced trade-off between accuracy and efficiency.

cs.CV↗

arXiv2024

Causal Perception Inspired Representation Learning for Trustworthy Image Quality Assessment

Despite great success in modeling visual perception, deep neural network based image quality assessment (IQA) still remains unreliable in real-world applications due to its vulnerability to adversarial perturbations and the inexplicit black-box structure. In this paper, we propose to build a trustworthy IQA model via Causal Perception inspired Representation Learning (CPRL), and a score reflection attack method for IQA model. More specifically, we assume that each image is composed of Causal Perception Representation (CPR) and non-causal perception representation (N-CPR). CPR serves as the causation of the subjective quality label, which is invariant to the imperceptible adversarial perturbations. Inversely, N-CPR presents spurious associations with the subjective quality label, which may significantly change with the adversarial perturbations. To extract the CPR from each input image, we develop a soft ranking based channel-wise activation function to mediate the causally sufficient (beneficial for high prediction accuracy) and necessary (beneficial for high robustness) deep features, and based on intervention employ minimax game to optimize. Experiments on four benchmark databases show that the proposed CPRL method outperforms many state-of-the-art adversarial defense methods and provides explicit model interpretation.

cs.CV↗

arXiv2024

Perceptual Constancy Constrained Single Opinion Score Calibration for Image Quality Assessment

In this paper, we propose a highly efficient method to estimate an image's mean opinion score (MOS) from a single opinion score (SOS). Assuming that each SOS is the observed sample of a normal distribution and the MOS is its unknown expectation, the MOS inference is formulated as a maximum likelihood estimation problem, where the perceptual correlation of pairwise images is considered in modeling the likelihood of SOS. More specifically, by means of the quality-aware representations learned from the self-supervised backbone, we introduce a learnable relative quality measure to predict the MOS difference between two images. Then, the current image's maximum likelihood estimation towards MOS is represented by the sum of another reference image's estimated MOS and their relative quality. Ideally, no matter which image is selected as the reference, the MOS of the current image should remain unchanged, which is termed perceptual cons tancy constrained calibration (PC3). Finally, we alternatively optimize the relative quality measure's parameter and the current image's estimated MOS via backpropagation and Newton's method respectively. Experiments show that the proposed method is efficient in calibrating the biased SOS and significantly improves IQA model learning when only SOSs are available.

cs.CV↗

arXiv2024

Beyond MOS: Subjective Image Quality Score Preprocessing Method Based on Perceptual Similarity

Image quality assessment often relies on raw opinion scores provided by subjects in subjective experiments, which can be noisy and unreliable. To address this issue, postprocessing procedures such as ITU-R BT.500, ITU-T P.910, and ITU-T P.913 have been standardized to clean up the original opinion scores. These methods use annotator-based statistical priors, but they do not take into account extensive information about the image itself, which limits their performance in less annotated scenarios. Generally speaking, image quality datasets usually contain similar scenes or distortions, and it is inevitable for subjects to compare images to score a reasonable score when scoring. Therefore, In this paper, we proposed Subjective Image Quality Score Preprocessing Method perceptual similarity Subjective Preprocessing (PSP), which exploit the perceptual similarity between images to alleviate subjective bias in less annotated scenarios. Specifically, we model subjective scoring as a conditional probability model based on perceptual similarity with previously scored images, called subconscious reference scoring. The reference images are stored by a neighbor dictionary, which is obtained by a normalized vector dot-product based nearest neighbor search of the images' perceptual depth features. Then the preprocessed score is updated by the exponential moving average (EMA) of the subconscious reference scoring, called similarity regularized EMA. Our experiments on multiple datasets (LIVE, TID2013, CID2013) show that this method can effectively remove the bias of the subjective scores. Additionally, Experiments prove that the Preprocesed dataset can improve the performance of downstream IQA tasks very well.

cs.CV↗

arXiv2024

Blood Works for Graphene Production

Blood, a ubiquitous and fundamental carbohydrate material composed of plasma, red blood cells, white blood cells, and platelets, has been playing an important role in biology, life science, history, and religious study, while graphene has garnered significant attention due to its exceptional properties and extensive range of potential applications. Achieving environmentally friendly, cost-effective growth using hybrid precursors and obtaining high-quality graphene through a straightforward CVD process has been traditionally considered mutually exclusive. This study demonstrates that we can produce high-quality graphene domains with controlled thickness through a one-step growth process at atmospheric pressure using blood as a precursor. Raman spectroscopy confirms the uniformity of the blood-grown graphene films, and observing the half-integer quantum Hall effect in the measured devices highlights its outstanding electronic properties. This unprecedented approach opens possibilities for blood application, facilitating an unconventional route in graphene growth applications.

cond-mat.mes-hall↗

arXiv2024

Gate control of 2D magnetism in tri- and four-layers $\rm CrI_3$/graphene heterostructures

We conduct experimental studies on the electrical transport properties of monolayer graphene directly covered by a few layers of $\rm CrI_3$. We do not observe the expected magnetic exchange coupling in the graphene but instead discover proximity effects featuring gate and magnetic field tunability. The tunability of gate voltage is manifested in the alignment of the lowest conduction band of $\rm CrI_3$ and the Fermi level of graphene, which can be controlled by the gate voltage. The coexistence of the normal and atypical quantum Hall effects in our device also corresponds to gate-control modulation doping. The lowest conduction band depends on the magnetic states of the $\rm CrI_3$ and can be altered by the magnetic field, which corresponds to the resistance loops during back-and-forth sweeps of the magnetic field. Our results serve as a reference for exploiting the magnetic proximity effects in graphene.

cond-mat.mes-hall↗

arXiv2024

Ferroelectricity in twisted double bilayer graphene

Two-dimensional ferroelectrics can maintain vertical polarization up to room temperature, and are, therefore, promising for next-generation nonvolatile memories. Although natural two-dimensional ferroelectrics are few, moiré superlattices provide us with a generalized method to construct ferroelectrics from non-ferroelectric parent materials. We report a realization of ferroelectric hysteresis in a AB-BA stacked twisted double bilayer graphene (TDBG) system. The ferroelectric polarization is prominent at zero external displacement field and reduces upon increasing displacement fields. TDBG in the AB-BA configuration possesses a superlattice of non-centrosymmetric domains, exhibiting alternatively switchable polarities even without the assistance of any boron nitride layers; however, in the AB-AB stacking case, the development of polarized domains necessitates the presence of a second superlattice induced by the adjacent boron nitride layer. Therefore, twisted multilayer graphene systems offer us a fascinating field to explore two-dimensional ferroelectricity.

cond-mat.mes-hall↗

arXiv2024

Tailoring coercive fields and the Curie temperature via proximity coupling in WSe$_2$/Fe$_3$GeTe$_2$ van der Waals heterostructures

Hybrid structures consisting of two-dimensional (2D) magnets and semiconductors have exhibited extensive functionalities in spintronics and opto-spintronics. In this work, we have fabricated WSe$_2$/Fe$_3$GeTe$_2$ van der Waals (vdW) heterostructures and investigated the proximity effects on 2D magnetism. Through reflective magnetic circular dichroism (RMCD), we have observed a temperature-dependent modulation of magnetic order in the heterostructure. For temperatures above $40$ K, WSe$_2$-covered Fe$_3$GeTe$_2$ exhibits a larger coercive field than that observed in bare Fe$_3$GeTe$_2$, accompanied by a noticeable enhancement of the Curie temperature by $21$ K. This strengthening suggests an increase in magnetic anisotropy in the interfacial Fe$_3$GeTe$_2$ layer, which can be attributed to the spin-orbit coupling (SOC) proximity effect induced by the adjacent WSe$_2$ layers. However, at much lower temperatures ($T<20$ K), a non-monotonic modification of the coercive field is observed, showing both reduction and enhancement, which depends on the thickness of the WSe$_2$ and Fe$_3$GeTe$_2$ layers. Moreover, an unconventional two-step magnetization process emerges in the heterostructure, indicating the short-range nature of SOC proximity effects. Our findings revealing proximity effects on 2D magnetism may shed light on the design of future spintronic and memory devices based on 2D magnetic heterostructures.

cond-mat.mes-hall↗

arXiv2024

MRScore: Evaluating Radiology Report Generation with LLM-based Reward System

In recent years, automated radiology report generation has experienced significant growth. This paper introduces MRScore, an automatic evaluation metric tailored for radiology report generation by leveraging Large Language Models (LLMs). Conventional NLG (natural language generation) metrics like BLEU are inadequate for accurately assessing the generated radiology reports, as systematically demonstrated by our observations within this paper. To address this challenge, we collaborated with radiologists to develop a framework that guides LLMs for radiology report evaluation, ensuring alignment with human analysis. Our framework includes two key components: i) utilizing GPT to generate large amounts of training data, i.e., reports with different qualities, and ii) pairing GPT-generated reports as accepted and rejected samples and training LLMs to produce MRScore as the model reward. Our experiments demonstrate MRScore's higher correlation with human judgments and superior performance in model selection compared to traditional metrics. Our code and datasets will be available on GitHub.

cs.CL↗

arXiv2024

Unpaired Multi-view Clustering via Reliable View Guidance

This paper focuses on unpaired multi-view clustering (UMC), a challenging problem where paired observed samples are unavailable across multiple views. The goal is to perform effective joint clustering using the unpaired observed samples in all views. In incomplete multi-view clustering, existing methods typically rely on sample pairing between views to capture their complementary. However, that is not applicable in the case of UMC. Hence, we aim to extract the consistent cluster structure across views. In UMC, two challenging issues arise: uncertain cluster structure due to lack of label and uncertain pairing relationship due to absence of paired samples. We assume that the view with a good cluster structure is the reliable view, which acts as a supervisor to guide the clustering of the other views. With the guidance of reliable views, a more certain cluster structure of these views is obtained while achieving alignment between reliable views and other views. Then we propose Reliable view Guidance with one reliable view (RG-UMC) and multiple reliable views (RGs-UMC) for UMC. Specifically, we design alignment modules with one reliable view and multiple reliable views, respectively, to adaptively guide the optimization process. Also, we utilize the compactness module to enhance the relationship of samples within the same cluster. Meanwhile, an orthogonal constraint is applied to latent representation to obtain discriminate features. Extensive experiments show that both RG-UMC and RGs-UMC outperform the best state-of-the-art method by an average of 24.14\% and 29.42\% in NMI, respectively.

cs.CV↗

arXiv2024

Beyond ESM2: Graph-Enhanced Protein Sequence Modeling with Efficient Clustering

Proteins are essential to life's processes, underpinning evolution and diversity. Advances in sequencing technology have revealed millions of proteins, underscoring the need for sophisticated pre-trained protein models for biological analysis and AI development. Facebook's ESM2, the most advanced protein language model to date, leverages a masked prediction task for unsupervised learning, crafting amino acid representations with notable biochemical accuracy. Yet, it lacks in delivering functional protein insights, signaling an opportunity for enhancing representation quality.Our study addresses this gap by incorporating protein family classification into ESM2's training.This approach, augmented with Community Propagation-Based Clustering Algorithm, improves global protein representations, while a contextual prediction task fine-tunes local amino acid accuracy. Significantly, our model achieved state-of-the-art results in several downstream experiments, demonstrating the power of combining global and local methodologies to substantially boost protein representation quality.

q-bio.BM↗

arXiv2024

Adaptive Hybrid Masking Strategy for Privacy-Preserving Face Recognition Against Model Inversion Attack

The utilization of personal sensitive data in training face recognition (FR) models poses significant privacy concerns, as adversaries can employ model inversion attacks (MIA) to infer the original training data. Existing defense methods, such as data augmentation and differential privacy, have been employed to mitigate this issue. However, these methods often fail to strike an optimal balance between privacy and accuracy. To address this limitation, this paper introduces an adaptive hybrid masking algorithm against MIA. Specifically, face images are masked in the frequency domain using an adaptive MixUp strategy. Unlike the traditional MixUp algorithm, which is predominantly used for data augmentation, our modified approach incorporates frequency domain mixing. Previous studies have shown that increasing the number of images mixed in MixUp can enhance privacy preservation but at the expense of reduced face recognition accuracy. To overcome this trade-off, we develop an enhanced adaptive MixUp strategy based on reinforcement learning, which enables us to mix a larger number of images while maintaining satisfactory recognition accuracy. To optimize privacy protection, we propose maximizing the reward function (i.e., the loss function of the FR system) during the training of the strategy network. While the loss function of the FR network is minimized in the phase of training the FR network. The strategy network and the face recognition network can be viewed as antagonistic entities in the training process, ultimately reaching a more balanced trade-off. Experimental results demonstrate that our proposed hybrid masking scheme outperforms existing defense algorithms in terms of privacy preservation and recognition accuracy against MIA.

cs.CV↗

arXiv2024

Droplet impact behavior on a hydrophobic plate with a wettability-patterned orifice

Droplet impact behavior has attracted much attention recently due to its academic significance and diverse industrial applications. This study employs the lattice Boltzmann method to simulate the impact of a droplet on a hydrophobic plate featuring a square orifice. Unlike previous studies, the chemical property of the orifice considered in this work is not homogeneous but heterogeneous, and its cross-sectional wettability changes from hydrophobicity to hydrophilicity. The study first validates the numerical method against experimental data, and then investigates in detail the influences of the Weber number, wettability difference, and pore size. According to the numerical results, we observed that the evolutionary stages of the impinging droplet always include the spreading phase and the rebounding phase, while whether there exists the splitting phase, it depends on the combined effect of the wettability difference and the Weber number. The impact behavior of droplets is analyzed by evaluating the underlying mechanisms such as kinetic energy, surface energy, viscous dissipation energy, and pressure. It is interesting to note that the existence of wettability-patterned pore tends to promote adhesion of droplets on the plate, resulting in the droplet impact behaviors are largely different from that for the case of homogeneous pore. Additionally, a phase diagram is constructed for various Weber numbers and pore sizes, revealing that the dynamic behavior of droplets is determined by the competition among dynamic pressure, capillary pressure, and viscous pressure losses. These insights from numerical studies guide the development of innovative solid substrates capable of manipulating droplet motion.

physics.flu-dyn↗