Search

Fully test-time adaptation aims to adapt a network model online based on sequential analysis of input samples during the inference stage. We observe that, when applying a transformer network model into a new domain, the self-attention profiles of image samples in the target domain deviate significantly from those in the source domain, which results in large performance degradation during domain changes. To address this important issue, we propose a new structure for the self-attention modules in the transformer. Specifically, we incorporate three domain-conditioning vectors, called domain conditioners, into the query, key, and value components of the self-attention module. We learn a network to generate these three domain conditioners from the class token at each transformer network layer. We find that, during fully online test-time adaptation, these domain conditioners at each transform network layer are able to gradually remove the impact of domain shift and largely recover the original self-attention profile. Our extensive experimental results demonstrate that the proposed domain-conditioned transformer significantly improves the online fully test-time domain adaptation performance and outperforms existing state-of-the-art methods by large margins.

Cloth-changing person re-identification (CC-ReID) poses a significant challenge in computer vision. A prevailing approach is to prompt models to concentrate on causal attributes, like facial features and hairstyles, rather than confounding elements such as clothing appearance. Traditional methods to achieve this involve integrating multi-modality data or employing manually annotated clothing labels, which tend to complicate the model and require extensive human effort. In our study, we demonstrate that simply reducing feature correlations during training can significantly enhance the baseline model's performance. We theoretically elucidate this effect and introduce a novel regularization technique based on density ratio estimation. This technique aims to minimize feature correlation in the training process of cloth-changing ReID baselines. Our approach is model-independent, offering broad enhancements without needing additional data or labels. We validate our method through comprehensive experiments on prevalent CC-ReID datasets, showing its effectiveness in improving baseline models' generalization capabilities.

Spatiotemporal neural networks have shown great promise in urban scenarios by effectively capturing temporal and spatial correlations. However, urban environments are constantly evolving, and current model evaluations are often limited to traffic scenarios and use data mainly collected only a few weeks after training period to evaluate model performance. The generalization ability of these models remains largely unexplored. To address this, we propose a Spatiotemporal Out-of-Distribution (ST-OOD) benchmark, which comprises six urban scenario: bike-sharing, 311 services, pedestrian counts, traffic speed, traffic flow, ride-hailing demand, and bike-sharing, each with in-distribution (same year) and out-of-distribution (next years) settings. We extensively evaluate state-of-the-art spatiotemporal models and find that their performance degrades significantly in out-of-distribution settings, with most models performing even worse than a simple Multi-Layer Perceptron (MLP). Our findings suggest that current leading methods tend to over-rely on parameters to overfit training data, which may lead to good performance on in-distribution data but often results in poor generalization. We also investigated whether dropout could mitigate the negative effects of overfitting. Our results showed that a slight dropout rate could significantly improve generalization performance on most datasets, with minimal impact on in-distribution performance. However, balancing in-distribution and out-of-distribution performance remains a challenging problem. We hope that the proposed benchmark will encourage further research on this critical issue.

Cloth-Changing Person Re-Identification (CC-ReID) involves recognizing individuals in images regardless of clothing status. In this paper, we empirically and experimentally demonstrate that completely eliminating or fully retaining clothing features is detrimental to the task. Existing work, either relying on clothing labels, silhouettes, or other auxiliary data, fundamentally aim to balance the learning of clothing and identity features. However, we practically find that achieving this balance is challenging and nuanced. In this study, we introduce a novel module called Diverse Norm, which expands personal features into orthogonal spaces and employs channel attention to separate clothing and identity features. A sample re-weighting optimization strategy is also introduced to guarantee the opposite optimization direction. Diverse Norm presents a simple yet effective approach that does not require additional data. Furthermore, Diverse Norm can be seamlessly integrated ResNet50 and significantly outperforms the state-of-the-art methods.

Traffic forecasting is a cornerstone of smart city management, enabling efficient resource allocation and transportation planning. Deep learning, with its ability to capture complex nonlinear patterns in spatiotemporal (ST) data, has emerged as a powerful tool for traffic forecasting. While graph neural networks (GCNs) and transformer-based models have shown promise, their computational demands often hinder their application to real-world road networks, particularly those with large-scale spatiotemporal interactions. To address these challenges, we propose a novel spatiotemporal graph transformer (STGformer) architecture. STGformer effectively balances the strengths of GCNs and Transformers, enabling efficient modeling of both global and local traffic patterns while maintaining a manageable computational footprint. Unlike traditional approaches that require multiple attention layers, STG attention block captures high-order spatiotemporal interactions in a single layer, significantly reducing computational cost. In particular, STGformer achieves a 100x speedup and a 99.8\% reduction in GPU memory usage compared to STAEformer during batch inference on a California road graph with 8,600 sensors. We evaluate STGformer on the LargeST benchmark and demonstrate its superiority over state-of-the-art Transformer-based methods such as PDFormer and STAEformer, which underline STGformer's potential to revolutionize traffic forecasting by overcoming the computational and memory limitations of existing approaches, making it a promising foundation for future spatiotemporal modeling tasks.

Recent advancements in Spatiotemporal Graph Neural Networks (ST-GNNs) and Transformers have demonstrated promising potential for traffic forecasting by effectively capturing both temporal and spatial correlations. The generalization ability of spatiotemporal models has received considerable attention in recent scholarly discourse. However, no substantive datasets specifically addressing traffic out-of-distribution (OOD) scenarios have been proposed. Existing ST-OOD methods are either constrained to testing on extant data or necessitate manual modifications to the dataset. Consequently, the generalization capacity of current spatiotemporal models in OOD scenarios remains largely underexplored. In this paper, we investigate state-of-the-art models using newly proposed traffic OOD benchmarks and, surprisingly, find that these models experience a significant decline in performance. Through meticulous analysis, we attribute this decline to the models' inability to adapt to previously unobserved spatial relationships. To address this challenge, we propose a novel Mixture of Experts (MoE) framework, which learns a set of graph generators (i.e., graphons) during training and adaptively combines them to generate new graphs based on novel environmental conditions to handle spatial distribution shifts during testing. We further extend this concept to the Transformer architecture, achieving substantial improvements. Our method is both parsimonious and efficacious, and can be seamlessly integrated into any spatiotemporal model, outperforming current state-of-the-art approaches in addressing spatial dynamics.

In the era of social media video platforms, popular ``hot-comments'' play a crucial role in attracting user impressions of short-form videos, making them vital for marketing and branding purpose. However, existing research predominantly focuses on generating descriptive comments or ``danmaku'' in English, offering immediate reactions to specific video moments. Addressing this gap, our study introduces \textsc{HotVCom}, the largest Chinese video hot-comment dataset, comprising 94k diverse videos and 137 million comments. We also present the \texttt{ComHeat} framework, which synergistically integrates visual, auditory, and textual data to generate influential hot-comments on the Chinese video dataset. Empirical evaluations highlight the effectiveness of our framework, demonstrating its excellence on both the newly constructed and existing datasets., Comment: Accepted to ACL 2024 (Findings)

Recent studies have shown that Vision Language Large Models (VLLMs) may output content not relevant to the input images. This problem, called the hallucination phenomenon, undoubtedly degrades VLLM performance. Therefore, various anti-hallucination techniques have been proposed to make model output more reasonable and accurate. Despite their successes, from extensive tests we found that augmenting the prompt (e.g. word appending, rewriting, and spell error etc.) may change model output and make the output hallucinate again. To cure this drawback, we propose a new instruct-tuning framework called Prompt Augmentation and Caption Utilization (PACU) to boost VLLM's generation ability under the augmented prompt scenario. Concretely, on the one hand, PACU exploits existing LLMs to augment and evaluate diverse prompts automatically. The resulting high-quality prompts are utilized to enhance VLLM's ability to process different prompts. On the other hand, PACU exploits image captions to jointly work with image features as well as the prompts for response generation. When the visual feature is inaccurate, LLM can capture useful information from the image captions for response generation. Extensive experiments on hallucination evaluation and prompt-augmented datasets demonstrate that our PACU method can work well with existing schemes to effectively boost VLLM model performance. Code is available in https://github.com/zhaominyiz/PACU.

The drive to control tokamaks, a prominent technology in nuclear fusion, is essential due to its potential to provide a virtually unlimited source of clean energy. Reinforcement learning (RL) promises improved flexibility to manage the intricate and non-linear dynamics of the plasma encapsulated in a tokamak. However, RL typically requires substantial interaction with a simulator capable of accurately evolving the high-dimensional plasma state. Compared to first-principle-based simulators, whose intense computations lead to sluggish RL training, we devise an effective method to acquire a fully data-driven simulator, by mitigating the arising compounding error issue due to the underlying autoregressive nature. With high accuracy and appealing extrapolation capability, this high-fidelity dynamics model subsequently enables the rapid training of a qualified RL agent to directly generate engineering-reasonable magnetic coil commands, aiming at the desired long-term targets of plasma current and last closed flux surface. Together with a surrogate magnetic equilibrium reconstruction model EFITNN, the RL agent successfully maintains a $100$-ms, $1$ kHz trajectory control with accurate waveform tracking on the HL-3 tokamak. Furthermore, it also demonstrates the feasibility of zero-shot adaptation to changed triangularity targets, confirming the robustness of the developed data-driven dynamics model. Our work underscores the advantage of fully data-driven dynamics models in yielding RL-based trajectory control policies at a sufficiently fast pace, an anticipated engineering requirement in daily discharge practices for the upcoming ITER device.

Diffusion models have achieved remarkable success in sequential decision-making by leveraging the highly expressive model capabilities in policy learning. A central problem for learning diffusion policies is to align the policy output with human intents in various tasks. To achieve this, previous methods conduct return-conditioned policy generation or Reinforcement Learning (RL)-based policy optimization, while they both rely on pre-defined reward functions. In this work, we propose a novel framework, Forward KL regularized Preference optimization for aligning Diffusion policies, to align the diffusion policy with preferences directly. We first train a diffusion policy from the offline dataset without considering the preference, and then align the policy to the preference data via direct preference optimization. During the alignment phase, we formulate direct preference learning in a diffusion policy, where the forward KL regularization is employed in preference optimization to avoid generating out-of-distribution actions. We conduct extensive experiments for MetaWorld manipulation and D4RL tasks. The results show our method exhibits superior alignment with preferences and outperforms previous state-of-the-art algorithms.

Vertical federated learning is a natural and elegant approach to integrate multi-view data vertically partitioned across devices (clients) while preserving their privacies. Apart from the model training, existing methods requires the collaboration of all clients in the model inference. However, the model inference is probably maintained for service in a long time, while the collaboration, especially when the clients belong to different organizations, is unpredictable in real-world scenarios, such as concellation of contract, network unavailablity, etc., resulting in the failure of them. To address this issue, we, at the first attempt, propose a flexible Active-Passive Federated learning (APFed) framework. Specifically, the active client is the initiator of a learning task and responsible to build the complete model, while the passive clients only serve as assistants. Once the model built, the active client can make inference independently. In addition, we instance the APFed framework into two classification methods with employing the reconstruction loss and the contrastive loss on passive clients, respectively. Meanwhile, the two methods are tested in a set of experiments and achieves desired results, validating their effectiveness.

We propose a group information geometry approach (GIGA) for ultra-massive multiple-input multiple-output (MIMO) signal detection. The signal detection task is framed as computing the approximate marginals of the a posteriori distribution of the transmitted data symbols of all users. With the approximate marginals, we perform the maximization of the {\textsl{a posteriori}} marginals (MPM) detection to recover the symbol of each user. Based on the information geometry theory and the grouping of the components of the received signal, three types of manifolds are constructed and the approximate a posteriori marginals are obtained through m-projections. The Berry-Esseen theorem is introduced to offer an approximate calculation of the m-projection, while its direct calculation is exponentially complex. In most cases, more groups, less complexity of GIGA. However, when the number of groups exceeds a certain threshold, the complexity of GIGA starts to increase. Simulation results confirm that the proposed GIGA achieves better bit error rate (BER) performance within a small number of iterations, which demonstrates that it can serve as an efficient detection method in ultra-massive MIMO systems.

Nonlinear circuits serve as crucial carriers and physical models for investigating nonlinear dynamics and chaotic behavior, particularly in the simulation of biological neurons. In this study, Chua's circuit and Lorenz circuit are systematically explored for the first time through machine learning correlation algorithms. Specifically, the upgraded and optimized SINDy-PI model, which is based on neural network and symbolic regression algorithm, is utilized to learn the numerical results of attractors generated by these two nonlinear circuits. Through error analysis, we examine the effects of the precision of input data and the amount of data on the algorithmic model. The findings reveal that when the input data quantity and data precision fall within a certain range, the algorithm model can effectively recognize and restore the differential equation expressions corresponding to the two circuits. Additionally, we test the anti-interference ability of different circuits and the robustness of the algorithm by introducing noise into the test data. The results indicate that under the same noise disturbance, the Lorenz circuit has better noise resistance than Chua's circuit, providing a starting point for further studying the intrinsic properties and characteristics of different nonlinear circuits. The above results will not only offer a reference for the further study of nonlinear circuits and related systems using deep learning algorithms but also lay a preliminary theoretical foundation for the study of related physical problems and applications., Comment: 13 pages, 7 figures, 4 tables

Continuous electroencephalography (EEG) signals are widely used in affective brain-computer interface (aBCI) applications. However, not all continuously collected EEG signals are relevant or meaningful to the task at hand (e.g., wondering thoughts). On the other hand, manually labeling the relevant parts is nearly impossible due to varying engagement patterns across different tasks and individuals. Therefore, effectively and efficiently identifying the important parts from continuous EEG recordings is crucial for downstream BCI tasks, as it directly impacts the accuracy and reliability of the results. In this paper, we propose a novel unsupervised deep reinforcement learning framework, called Emotion-Agent, to automatically identify relevant and informative emotional moments from continuous EEG signals. Specifically, Emotion-Agent involves unsupervised deep reinforcement learning combined with a heuristic algorithm. We first use the heuristic algorithm to perform an initial global search and form prototype representations of the EEG signals, which facilitates the efficient exploration of the signal space and identify potential regions of interest. Then, we design distribution-prototype reward functions to estimate the interactions between samples and prototypes, ensuring that the identified parts are both relevant and representative of the underlying emotional states. Emotion-Agent is trained using Proximal Policy Optimization (PPO) to achieve stable and efficient convergence. Our experiments compare the performance with and without Emotion-Agent. The results demonstrate that selecting relevant and informative emotional parts before inputting them into downstream tasks enhances the accuracy and reliability of aBCI applications., Comment: 11 pages, 4 figures, 4 tables, submitted to AAAI 2025

Many dipolar topological structures have been experimentally demonstrated in (PbTiO$_3$)$_n$/(SrTiO$_3$)$_n$ superlattices, such as flux-closure, vortice, and skyrmion. In this work, we employ deep potential molecular dynamics (MD) to investigate the dynamical response of the (PbTiO$_3$)$_{10}$/(SrTiO$_3$)$_{10}$ superlattice, supporting polar vortex arrays, to temperature and electric field at the atomic level. Our simulations reveal a unique phase transition sequence from ferroelectric-like to antiferroelectric-like to paraelectric in the in-plane direction as temperature increases. In the ferroelectric-like state, we observe field-driven reversible switching of in-plane polarization coupled with out-of-plane movements of vortex cores during MD simulations. In the antiferroelectric-like region, the polarization-electric field hysteresis loop exhibits a superparaelectric feature, showing nearly no loss. This behavior is attributed to a strong recovering force to form polar vortex arrays, dictated by the electrical and mechanical boundary conditions within the superlattice. The (PbTiO$_3$)$_{10}$/(SrTiO$_3$)$_{10}$ superlattice in the antiferroelectric-like state also demonstrates large in-plane susceptibility and tunability. The effect of Pb doping in the SrTiO$_3$ layer on the topological structural transition in the superlattice is investigated. The weakened depolarization field in the PbTiO$_3$ layers leads to new dipolar configurations, such as enlarged skyrmion bubble within $c$ domains in (PbTiO$_3$)$_{10}$/(Pb$_{0.4}$Sr$_{0.6}$TiO$_3$)$_{10}$, and we quantify their thermal and electrical responses through MD simulations. These quantitative atomistic insights will be useful for the controlled optimization of perovskite superlattices for various device applications., Comment: 8 figures

In this paper, we propose an information geometry (IG) framework to solve the standard linear regression problem. The proposed framework is an extension of the one for computing the mean of complex multivariate Gaussian distribution. By applying the proposed framework, the information geometry approach (IGA) and the approximate information geometry approach (AIGA) for basis pursuit de-noising (BPDN) in standard linear regression are derived. The framework can also be applied to other standard linear regression problems. With the transformations of natural and expectation parameters of Gaussian distributions, we then show the relationship between the IGA and the message passing (MP) algorithm. Finally, we prove that the AIGA is equivalent to the approximate message passing (AMP) algorithm. These intrinsic results offer a new perspective for the AMP algorithm, and clues for understanding and improving stochastic reasoning methods., Comment: 30 pages, 5 figures

This paper presents the design and optimization of a LYSO crystal electromagnetic calorimeter (ECAL) for the DarkSHINE experiment, which aims to search for dark photons as potential mediators of dark forces. The ECAL design was evaluated through comprehensive simulations, focusing on optimizing dimensions, material selection, energy distribution, and energy resolution. The ECAL configuration consists of 21$\times$21$\times$11 LYSO crystals, each measuring 2.5$\times$2.5$\times$4 cm$^3$, arranged in a staggered layout to improve signal detection efficiency. A 4 GeV energy dynamic range was established to ensure accurate energy measurements without saturation, which is essential for background rejection and signal identification. A detailed digitization model was developed to simulate the scintillation, SiPM, and ADC behaviors, providing a more realistic representation of detector performance. Additionally, the study assessed radiation damage in the ECAL region, highlighting the necessity of radiation-resistant scintillators and silicon sensors.

Vision foundation models are increasingly employed in autonomous driving systems due to their advanced capabilities. However, these models are susceptible to adversarial attacks, posing significant risks to the reliability and safety of autonomous vehicles. Adversaries can exploit these vulnerabilities to manipulate the vehicle's perception of its surroundings, leading to erroneous decisions and potentially catastrophic consequences. To address this challenge, we propose a novel Precision-Guided Adversarial Attack (PG-Attack) framework that combines two techniques: Precision Mask Perturbation Attack (PMP-Attack) and Deceptive Text Patch Attack (DTP-Attack). PMP-Attack precisely targets the attack region to minimize the overall perturbation while maximizing its impact on the target object's representation in the model's feature space. DTP-Attack introduces deceptive text patches that disrupt the model's understanding of the scene, further enhancing the attack's effectiveness. Our experiments demonstrate that PG-Attack successfully deceives a variety of advanced multi-modal large models, including GPT-4V, Qwen-VL, and imp-V1. Additionally, we won First-Place in the CVPR 2024 Workshop Challenge: Black-box Adversarial Attacks on Vision Foundation Models and codes are available at https://github.com/fuhaha824/PG-Attack., Comment: First-Place in the CVPR 2024 Workshop Challenge: Black-box Adversarial Attacks on Vision Foundation Models

Novel View Synthesis plays a crucial role by generating new 2D renderings from multi-view images of 3D scenes. However, capturing high-speed scenes with conventional cameras often leads to motion blur, hindering the effectiveness of 3D reconstruction. To address this challenge, high-frame-rate dense 3D reconstruction emerges as a vital technique, enabling detailed and accurate modeling of real-world objects or scenes in various fields, including Virtual Reality or embodied AI. Spike cameras, a novel type of neuromorphic sensor, continuously record scenes with an ultra-high temporal resolution, showing potential for accurate 3D reconstruction. Despite their promise, existing approaches, such as applying Neural Radiance Fields (NeRF) to spike cameras, encounter challenges due to the time-consuming rendering process. To address this issue, we make the first attempt to introduce the 3D Gaussian Splatting (3DGS) into spike cameras in high-speed capture, providing 3DGS as dense and continuous clues of views, then constructing SpikeGS. Specifically, to train SpikeGS, we establish computational equations between the rendering process of 3DGS and the processes of instantaneous imaging and exposing-like imaging of the continuous spike stream. Besides, we build a very lightweight but effective mapping process from spikes to instant images to support training. Furthermore, we introduced a new spike-based 3D rendering dataset for validation. Extensive experiments have demonstrated our method possesses the high quality of novel view rendering, proving the tremendous potential of spike cameras in modeling 3D scenes.

In this paper, we prove that on a smooth K\"ahler manifold, the $\mathbb{G}$-coercivity of the weighted Mabuchi functional implies the existence of the (v, w)-weighted-cscK (extremal) metric with v log-concave (firstly studied in \cite{Lah19}), e.g, cscK metrics, K\"ahler-Ricci solitons, $\mu$-cscK metrics., Comment: There is an error in formula (2.18) of the first version. This Error introduces a new difficulty in the W^2,p estimate, but does not influence the validity of other proofs. 1. We add the log-concavity assumption to overcome this difficulty. 2. We remove the result for the singular case, where we need to use Chern-Lu inequality instead, which produces a different sign for the error terms

Learning predictive world models is essential for enhancing the planning capabilities of reinforcement learning agents. Notably, the MuZero-style algorithms, based on the value equivalence principle and Monte Carlo Tree Search (MCTS), have achieved superhuman performance in various domains. However, in environments that require capturing long-term dependencies, MuZero's performance deteriorates rapidly. We identify that this is partially due to the \textit{entanglement} of latent representations with historical information, which results in incompatibility with the auxiliary self-supervised state regularization. To overcome this limitation, we present \textit{UniZero}, a novel approach that \textit{disentangles} latent states from implicit latent history using a transformer-based latent world model. By concurrently predicting latent dynamics and decision-oriented quantities conditioned on the learned latent history, UniZero enables joint optimization of the long-horizon world model and policy, facilitating broader and more efficient planning in latent space. We demonstrate that UniZero, even with single-frame inputs, matches or surpasses the performance of MuZero-style algorithms on the Atari 100k benchmark. Furthermore, it significantly outperforms prior baselines in benchmarks that require long-term memory. Lastly, we validate the effectiveness and scalability of our design choices through extensive ablation studies, visual analyses, and multi-task learning results. The code is available at \textcolor{magenta}{https://github.com/opendilab/LightZero}., Comment: 32 pages, 16 figures

Hundreds of 650 MHz superconducting radio-frequency (SRF) cavities with high intrinsic quality factor (Q0) and accelerating gradient (Eacc) will be adopted for Circular Electron Positron Collider (CEPC). The values of Q0 and Eacc are obtained during vertical test at 2.0 K. Hence, high accuracy of vertical test is essential for evaluating the performance of SRF cavity. The 650 MHz SRF cavities achieved very high Q0 (6E10) and Eacc (40 MV/m) during the vertical test. In our study, the error analysis of vertical test was conducted in the scalar case, in order to achieve high accuracy. The uncertainties of vertical test were obtained through calculation, which was approximately 3% for Eacc and less than 5% for Q0. This result was reasonable and acceptable.

The amplification of high-speed micro-motions holds significant promise, with applications spanning fault detection in fast-paced industrial environments to refining precision in medical procedures. However, conventional motion magnification algorithms often encounter challenges in high-speed scenarios due to low sampling rates or motion blur. In recent years, spike cameras have emerged as a superior alternative for visual tasks in such environments, owing to their unique capability to capture temporal and spatial frequency domains with exceptional fidelity. Unlike conventional cameras, which operate at fixed, low frequencies, spike cameras emulate the functionality of the retina, asynchronously capturing photon changes at each pixel position using spike streams. This innovative approach comprehensively records temporal and spatial visual information, rendering it particularly suitable for magnifying high-speed micro-motions.This paper introduces SpikeMM, a pioneering spike-based algorithm tailored specifically for high-speed motion magnification. SpikeMM integrates multi-level information extraction, spatial upsampling, and motion magnification modules, offering a self-supervised approach adaptable to a wide range of scenarios. Notably, SpikeMM facilitates seamless integration with high-performance super-resolution and motion magnification algorithms. We substantiate the efficacy of SpikeMM through rigorous validation using scenes captured by spike cameras, showcasing its capacity to magnify motions in real-world high-frequency settings.

Recent progress in Neural Causal Models (NCMs) showcased how identification and partial identification of causal effects can be automatically carried out via training of neural generative models that respect the constraints encoded in a given causal graph [Xia et al. 2022, Balazadeh et al. 2022]. However, formal consistency of these methods has only been proven for the case of discrete variables or only for linear causal models. In this work, we prove consistency of partial identification via NCMs in a general setting with both continuous and categorical variables. Further, our results highlight the impact of the design of the underlying neural network architecture in terms of depth and connectivity as well as the importance of applying Lipschitz regularization in the training phase. In particular, we provide a counterexample showing that without Lipschitz regularization the NCM may not be asymptotically consistent. Our results are enabled by new results on the approximability of structural causal models via neural generative models, together with an analysis of the sample complexity of the resulting architectures and how that translates into an error in the constrained optimization problem that defines the partial identification bounds., Comment: 37 pages, 8 figures

Recently, a new kind of collinear magnetism, dubbed altermagnetism, has attracted considerable interests. A key characteristic of altermagnet is the momentum-dependent band and spin splitting without net magnetization. However, finding altermagnetic materials with large splitting near the Fermi level, which necessarily requires three-dimensional k-space mapping and is crucial for spintronic applications and emergent phenomena, remains challenging. Here by employing synchrotron-based angle-resolved photoemission spectroscopy (ARPES), spin-resolved ARPES and model calculations, we uncover a large altermagnetic splitting, up to ~1.0 eV, near the Fermi level in CrSb. We verify its bulk-type g-wave altermagnetism through systematic three-dimensional k-space mapping, which unambiguously reveals the altermagnetic symmetry and associated nodal planes. The ARPES results are well captured by density functional theory calculations. Spin-resolved ARPES measurements further verify the spin polarizations of the split bands near Fermi level. In addition, tight-binding model analysis indicates that the large altermagnetic splitting arises from strong third-nearest-neighbor hopping mediated by Sb ions, which breaks both the space-time reversal symmetry and the translational spin-rotation symmetry. The large band/spin splitting near Fermi level in metallic CrSb, together with its high TN (up to 705 K) and simple spin configuration, paves the way for exploring emergent phenomena and spintronic applications based on altermagnets., Comment: The manuscript contains 19 pages, 5 figures and 1 table. The supplementary information contains 15 pages, 13 figures and 5 tables

In-situ measurements of the solar wind, a turbulent and anisotropic plasma flow originating at the Sun, are mostly carried out by single spacecraft, resulting in one-dimensional time series. The conversion of these measurements to the spatial frame of the plasma is a great challenge, but required for direct comparison of the measurements with MHD turbulence theories. Here we present a toolkit, based on the synthetic modeling of solar wind fluctuations as two-dimensional noise maps with adjustable spectral and power anisotropy, that can help with the temporal-spatial conversion of real data. Specifically, by following the spacecraft trajectory through a noise map (relative velocity and angle relative to some mean magnetic field) with properties tuned to mimic those of the solar wind, the likelihood that the temporal data fluctuations represent parallel or perpendicular fluctuations in the plasma frame can be quantified by correlating structure functions of the noise map. Synthetic temporal data can also be generated, which can provide a testing ground for analysis applied to the solar wind data. We demonstrate this tool by investigating Parker Solar Probe's E7 trajectory and data, and showcase several possible ways in which it can be used. We find that whether temporal variations in the spacecraft frame come from parallel or perpendicular variations in the plasma frame strongly depends on the spectral and power anisotropy of the measured wind. Data analysis assisted by such underlying synthetic models as presented here could open up new ways to interpret measurements in the future, specifically in the more reliable determination of plasma frame quantities from temporal measurements., Comment: Accepted as of 19/05/2024

We study the ground-state properties of dipolar spin-1/2 Bose-Einstein condensates with quantum fluctuations and Rashba spin-orbit coupling (SOC). The combined effects of dipole-dipole interaction (DDI), SOC, and Lee-Huang-Yang (LHY) correction induced by quantum fluctuations on the ground-state structures and spin textures of the system are analyzed and discussed. For the nonrotating case and fixed nonlinear interspecies contact interaction strengths, our results show that structural phase transitions can be achieved by adjusting the strengths of the DDI and LHY correction. In the absence of SOC, a ground-state phase diagram is given with respect to the DDI strength and the LHY correction strength. We find that the system exhibits rich quantum phases including square droplet lattice phase, annular phase, loop-island structure, stripe-droplet coexistence phase, toroidal stripe phase, and Thomas-Fermi (TF) phase. For the rotating case, the increase of DDI strength can lead to a quantum phase transition from superfluid phase to supersolid phase. In the presence of SOC, the quantum droplets display obvious stretching and hidden vortex-antivortex clusters are formed in each component. In particular, weak or moderate SOC favors the formation of droplets while for strong SOC the ground state of the system develops into a stripe phase with hidden vortex-antivortex clusters. Furthermore, the system sustains exotic spin textures and topological excitations, such as composite skyrmion-antiskyrmion-meron-antimeron cluster, meron-antimeron string cluster, antimeron-meron-antimeron chain cluster, and peculiar skyrmion-antiskyrmion-meron-antimeron necklace with a meron-antimeron necklace embedded inside and a central spin Neel domain wall., Comment: 11 pages,7 figures, accepted for publication in Chaos, Solitons and Fractals

Compiler technologies in deep learning and domain-specific hardware acceleration are increasingly adopting extensible compiler frameworks such as Multi-Level Intermediate Representation (MLIR) to facilitate more efficient development. With MLIR, compiler developers can easily define their own custom IRs in the form of MLIR dialects. However, the diversity and rapid evolution of such custom IRs make it impractical to manually write a custom test generator for each dialect. To address this problem, we design a new test generator called SYNTHFUZZ that combines grammar-based fuzzing with custom mutation synthesis. The key essence of SYNTHFUZZ is two fold: (1) It automatically infers parameterized context-dependent custom mutations from existing test cases. (2) It then concretizes the mutation's content depending on the target context and reduces the chance of inserting invalid edits by performing k-ancestor and pre(post)fix matching. SYNTHFUZZ obviates the need to manually define custom mutation operators for each dialect. We compare SYNTHFUZZ to three baselines: Grammarinator, MLIRSmith, and NeuRI. We conduct this comprehensive comparison on four different MLIR projects. Each project defines a new set of MLIR dialects where manually writing a custom test generator would take weeks of effort. Our evaluation shows that SYNTHFUZZ on average improves MLIR dialect pair coverage by 1.75 times, which increases branch coverage by 1.22 times. Further, we show that our context dependent custom mutation increases the proportion of valid tests by up to 1.11 times, indicating that SYNTHFUZZ correctly concretizes its parameterized mutations with respect to the target context. Parameterization of the mutations reduces the fraction of tests violating the base MLIR constraints by 0.57 times, increasing the time spent fuzzing dialect-specific code.

The gravitational microlensing method of discovering exoplanets and multi-star systems can produce degenerate solutions, some of which require in-depth analysis to uncover. We propose a new parameter space that can be used to sample potential solutions more efficiently and is more robust at finding all degenerate solutions. We identified two new parameters, k and h, that can be sampled in place of the mass ratios and separations of the systems under analysis to identify degenerate solutions. The parameter k is related to the size of the central caustic, $\Delta\xi_c$, while h is related to the distance of a point along the k contour from log(s)=0, where s is the projected planet-host separation. In this work, we present the characteristics of these parameters and the tests we conducted to prove their efficacy., Comment: 18 pages, updated with version accepted to PASP

Global wildfire spreading dynamics and severity are analyzed using the susceptible-infected-recovered (SIR) compartment model. We use the novel FireTracks (FT) Scientific Dataset covering the wildfire time series of 2002-2023.

Recently, diffusion-based depth estimation methods have drawn widespread attention due to their elegant denoising patterns and promising performance. However, they are typically unreliable under adverse conditions prevalent in real-world scenarios, such as rainy, snowy, etc. In this paper, we propose a novel robust depth estimation method called D4RD, featuring a custom contrastive learning mode tailored for diffusion models to mitigate performance degradation in complex environments. Concretely, we integrate the strength of knowledge distillation into contrastive learning, building the `trinity' contrastive scheme. This scheme utilizes the sampled noise of the forward diffusion process as a natural reference, guiding the predicted noise in diverse scenes toward a more stable and precise optimum. Moreover, we extend noise-level trinity to encompass more generic feature and image levels, establishing a multi-level contrast to distribute the burden of robust perception across the overall network. Before addressing complex scenarios, we enhance the stability of the baseline diffusion model with three straightforward yet effective improvements, which facilitate convergence and remove depth outliers. Extensive experiments demonstrate that D4RD surpasses existing state-of-the-art solutions on synthetic corruption datasets and real-world weather conditions. Source code and data are available at \url{https://github.com/wangjiyuan9/D4RD}., Comment: Accept by ACM MM 2024, Camera Ready Version

In recent years, Federated Unlearning (FU) has gained attention for addressing the removal of a client's influence from the global model in Federated Learning (FL) systems, thereby ensuring the ``right to be forgotten" (RTBF). State-of-the-art methods for unlearning use historical data from FL clients, such as gradients or locally trained models. However, studies have revealed significant information leakage in this setting, with the possibility of reconstructing a user's local data from their uploaded information. Addressing this, we propose Starfish, a privacy-preserving federated unlearning scheme using Two-Party Computation (2PC) techniques and shared historical client data between two non-colluding servers. Starfish builds upon existing FU methods to ensure privacy in unlearning processes. To enhance the efficiency of privacy-preserving FU evaluations, we suggest 2PC-friendly alternatives for certain FU algorithm operations. We also implement strategies to reduce costs associated with 2PC operations and lessen cumulative approximation errors. Moreover, we establish a theoretical bound for the difference between the unlearned global model via Starfish and a global model retrained from scratch for certified client removal. Our theoretical and experimental analyses demonstrate that Starfish achieves effective unlearning with reasonable efficiency, maintaining privacy and security in FL systems.

In many practical applications, usually, similar optimisation problems or scenarios repeatedly appear. Learning from previous problem-solving experiences can help adjust algorithm components of meta-heuristics, e.g., adaptively selecting promising search operators, to achieve better optimisation performance. However, those experiences obtained from previously solved problems, namely offline experiences, may sometimes provide misleading perceptions when solving a new problem, if the characteristics of previous problems and the new one are relatively different. Learning from online experiences obtained during the ongoing problem-solving process is more instructive but highly restricted by limited computational resources. This paper focuses on the effective combination of offline and online experiences. A novel hybrid framework that learns to dynamically and adaptively select promising search operators is proposed. Two adaptive operator selection modules with complementary paradigms cooperate in the framework to learn from offline and online experiences and make decisions. An adaptive decision policy is maintained to balance the use of those two modules in an online manner. Extensive experiments on 170 widely studied real-value benchmark optimisation problems and a benchmark set with 34 instances for combinatorial optimisation show that the proposed hybrid framework outperforms the state-of-the-art methods. Ablation study verifies the effectiveness of each component of the framework.

The integration of human emotions into multimedia applications shows great potential for enriching user experiences and enhancing engagement across various digital platforms. Unlike traditional methods such as questionnaires, facial expressions, and voice analysis, brain signals offer a more direct and objective understanding of emotional states. However, in the field of electroencephalography (EEG)-based emotion recognition, previous studies have primarily concentrated on training and testing EEG models within a single dataset, overlooking the variability across different datasets. This oversight leads to significant performance degradation when applying EEG models to cross-corpus scenarios. In this study, we propose a novel Joint Contrastive learning framework with Feature Alignment (JCFA) to address cross-corpus EEG-based emotion recognition. The JCFA model operates in two main stages. In the pre-training stage, a joint domain contrastive learning strategy is introduced to characterize generalizable time-frequency representations of EEG signals, without the use of labeled data. It extracts robust time-based and frequency-based embeddings for each EEG sample, and then aligns them within a shared latent time-frequency space. In the fine-tuning stage, JCFA is refined in conjunction with downstream tasks, where the structural connections among brain electrodes are considered. The model capability could be further enhanced for the application in emotion detection and interpretation. Extensive experimental results on two well-recognized emotional datasets show that the proposed JCFA model achieves state-of-the-art (SOTA) performance, outperforming the second-best method by an average accuracy increase of 4.09% in cross-corpus EEG-based emotion recognition tasks.

Fundus image segmentation on unseen domains is challenging, especially for the over-parameterized deep models trained on the small medical datasets. To address this challenge, we propose a method named Adaptive Feature-fusion Neural Network (AFNN) for glaucoma segmentation on unseen domains, which mainly consists of three modules: domain adaptor, feature-fusion network, and self-supervised multi-task learning. Specifically, the domain adaptor helps the pretrained-model fast adapt from other image domains to the medical fundus image domain. Feature-fusion network and self-supervised multi-task learning for the encoder and decoder are introduced to improve the domain generalization ability. In addition, we also design the weighted-dice-loss to improve model performance on complex optic-cup segmentation tasks. Our proposed method achieves a competitive performance over existing fundus segmentation methods on four public glaucoma datasets., Comment: 17 pages, 11 figures

Generative recommendation has emerged as a promising paradigm aimed at augmenting recommender systems with recent advancements in generative artificial intelligence. This task has been formulated as a sequence-to-sequence generation process, wherein the input sequence encompasses data pertaining to the user's previously interacted items, and the output sequence denotes the generative identifier for the suggested item. However, existing generative recommendation approaches still encounter challenges in (i) effectively integrating user-item collaborative signals and item content information within a unified generative framework, and (ii) executing an efficient alignment between content information and collaborative signals. In this paper, we introduce content-based collaborative generation for recommender systems, denoted as ColaRec. To capture collaborative signals, the generative item identifiers are derived from a pretrained collaborative filtering model, while the user is represented through the aggregation of interacted items' content. Subsequently, the aggregated textual description of items is fed into a language model to encapsulate content information. This integration enables ColaRec to amalgamate collaborative signals and content information within an end-to-end framework. Regarding the alignment, we propose an item indexing task to facilitate the mapping between the content-based semantic space and the interaction-based collaborative space. Additionally, a contrastive loss is introduced to ensure that items with similar collaborative GIDs possess comparable content representations, thereby enhancing alignment. To validate the efficacy of ColaRec, we conduct experiments on three benchmark datasets. Empirical results substantiate the superior performance of ColaRec.

Sequential Recommenders have been widely applied in various online services, aiming to model users' dynamic interests from their sequential interactions. With users increasingly engaging with online platforms, vast amounts of lifelong user behavioral sequences have been generated. However, existing sequential recommender models often struggle to handle such lifelong sequences. The primary challenges stem from computational complexity and the ability to capture long-range dependencies within the sequence. Recently, a state space model featuring a selective mechanism (i.e., Mamba) has emerged. In this work, we investigate the performance of Mamba for lifelong sequential recommendation (i.e., length>=2k). More specifically, we leverage the Mamba block to model lifelong user sequences selectively. We conduct extensive experiments to evaluate the performance of representative sequential recommendation models in the setting of lifelong sequences. Experiments on two real-world datasets demonstrate the superiority of Mamba. We found that RecMamba achieves performance comparable to the representative model while significantly reducing training duration by approximately 70% and memory costs by 80%. Codes and data are available at \url{https://github.com/nancheng58/RecMamba}.

In this paper we have studied the most general generating function of reduction for one loop integrals with arbitrary tensor structure in numerator and arbitrary power distribution of propagators in denominator. Using IBP relations, we have established the partial differential equations for these generating functions and solved them analytically. These results provide useful guidance for applying generating function method to reductions of higher loop integrals., Comment: 50 pages

Despite the substantial advancements in Vision-Language Pre-training (VLP) models, their susceptibility to adversarial attacks poses a significant challenge. Existing work rarely studies the transferability of attacks on VLP models, resulting in a substantial performance gap from white-box attacks. We observe that prior work overlooks the interaction mechanisms between modalities, which plays a crucial role in understanding the intricacies of VLP models. In response, we propose a novel attack, called Collaborative Multimodal Interaction Attack (CMI-Attack), leveraging modality interaction through embedding guidance and interaction enhancement. Specifically, attacking text at the embedding level while preserving semantics, as well as utilizing interaction image gradients to enhance constraints on perturbations of texts and images. Significantly, in the image-text retrieval task on Flickr30K dataset, CMI-Attack raises the transfer success rates from ALBEF to TCL, $\text{CLIP}_{\text{ViT}}$ and $\text{CLIP}_{\text{CNN}}$ by 8.11%-16.75% over state-of-the-art methods. Moreover, CMI-Attack also demonstrates superior performance in cross-task generalization scenarios. Our work addresses the underexplored realm of transfer attacks on VLP models, shedding light on the importance of modality interaction for enhanced adversarial robustness., Comment: This work won first place in CVPR 2024 Workshop Challenge: Black-box Adversarial Attacks on Vision Foundation Models

The high coercive field ($\mathcal{E}_c$) of hafnia-based ferroelectrics presents a major obstacle to their applications. The ferroelectric switching mechanisms in hafnia that dictate $\mathcal{E}_c$, especially those related to nucleation-and-growth at the domain wall (DW), have remained elusive. Through deep-learning-assisted multiscale simulations, we determine the finite-temperature thermodynamics and switching mechanisms for diverse types of 180$^\circ$ DWs, revealing a complex, stress-sensitive mobility landscape. The propagation velocities for mobile DW types under various thermal conditions can be characterized with a single creep equation, featuring a creep exponent of 2. This unconventional critical exponent results from the nucleation of a half-unit-cell-thin, elliptically-shaped critical nucleus. Our multiscale approach not only reproduces the experimental thickness ($d$) scaling, $\mathcal{E}_c\propto d^{-\frac{2}{3}}$, but also predicts that $\mathcal{E}_c$ of HfO$_2$ can be engineered to $\approx$0.1 MV/cm, even lower than perovskite ferroelectrics. The theoretical lower bound of $\mathcal{E}_c$ afforded by ferroelectric hafnia offers opportunities to realize power-efficient, high-fidelity ferroelectric nanoelectronics., Comment: 19 pages, 4 figures

Personalized federated learning becomes a hot research topic that can learn a personalized learning model for each client. Existing personalized federated learning models prefer to aggregate similar clients with similar data distribution to improve the performance of learning models. However, similaritybased personalized federated learning methods may exacerbate the class imbalanced problem. In this paper, we propose a novel Dynamic Affinity-based Personalized Federated Learning model (DA-PFL) to alleviate the class imbalanced problem during federated learning. Specifically, we build an affinity metric from a complementary perspective to guide which clients should be aggregated. Then we design a dynamic aggregation strategy to dynamically aggregate clients based on the affinity metric in each round to reduce the class imbalanced risk. Extensive experiments show that the proposed DA-PFL model can significantly improve the accuracy of each client in three real-world datasets with state-of-the-art comparison methods.

Reconstructing a sequence of sharp images from the blurry input is crucial for enhancing our insights into the captured scene and poses a significant challenge due to the limited temporal features embedded in the image. Spike cameras, sampling at rates up to 40,000 Hz, have proven effective in capturing motion features and beneficial for solving this ill-posed problem. Nonetheless, existing methods fall into the supervised learning paradigm, which suffers from notable performance degradation when applied to real-world scenarios that diverge from the synthetic training data domain. Moreover, the quality of reconstructed images is capped by the generated images based on motion analysis interpolation, which inherently differs from the actual scene, affecting the generalization ability of these methods in real high-speed scenarios. To address these challenges, we propose the first self-supervised framework for the task of spike-guided motion deblurring. Our approach begins with the formulation of a spike-guided deblurring model that explores the theoretical relationships among spike streams, blurry images, and their corresponding sharp sequences. We subsequently develop a self-supervised cascaded framework to alleviate the issues of spike noise and spatial-resolution mismatching encountered in the deblurring model. With knowledge distillation and re-blurring loss, we further design a lightweight deblur network to generate high-quality sequences with brightness and texture consistency with the original input. Quantitative and qualitative experiments conducted on our real-world and synthetic datasets with spikes validate the superior generalization of the proposed framework. Our code, data and trained models will be available at \url{https://github.com/chenkang455/S-SDM}., Comment: 14 pages

The rapid development of Large Language Models (LLMs) has facilitated a variety of applications from different domains. In this technical report, we explore the integration of LLMs and the popular academic writing tool, Overleaf, to enhance the efficiency and quality of academic writing. To achieve the above goal, there are three challenges: i) including seamless interaction between Overleaf and LLMs, ii) establishing reliable communication with the LLM provider, and iii) ensuring user privacy. To address these challenges, we present OverleafCopilot, the first-ever tool (i.e., a browser extension) that seamlessly integrates LLMs and Overleaf, enabling researchers to leverage the power of LLMs while writing papers. Specifically, we first propose an effective framework to bridge LLMs and Overleaf. Then, we developed PromptGenius, a website for researchers to easily find and share high-quality up-to-date prompts. Thirdly, we propose an agent command system to help researchers quickly build their customizable agents. OverleafCopilot (https://chromewebstore.google.com/detail/overleaf-copilot/eoadabdpninlhkkbhngoddfjianhlghb ) has been on the Chrome Extension Store, which now serves thousands of researchers. Additionally, the code of PromptGenius is released at https://github.com/wenhaomin/ChatGPT-PromptGenius. We believe our work has the potential to revolutionize academic writing practices, empowering researchers to produce higher-quality papers in less time.

Germanium arsenide (GeAs) is a layered semiconductor with remarkably anisotropic physical, thermoelectric and optical properties, and a promising candidate for multifunctional devices based on in-plane polarization dependent response. Understanding the underlying mechanism of such devices requires the knowledge of GeAs electronic band structure and of the hot carrier dynamics in its conduction band, whose details are still unclear. In this work, we investigated the properties of occupied and photoexcited states of GeAs in energy-momentum space, by combining scanning tunneling spectroscopy (STS), angle-resolved photoemission spectroscopy (ARPES) and time-resolved ARPES. We found that, GeAs is an indirect gap semiconductor having an electronic gap of 0.8 eV, for which the conduction band minimum (CBM) is located at the Gamma point while the valence band maximum (VBM) is out of Gamma. A Stark broadening of the valence band is observed immediately after photoexcitation, which can be attributed to the effects of the electrical field at the surface induced by inhomogeneous screening. Moreover, the hot electrons relaxation time of 1.56 ps down to the CBM which is dominated from both inter-valley and intra-valley coupling. Besides their relevance for our understanding of GeAs, these findings present general interest for the design on high performance thermoelectric and optoelectronic devices based on 2D semiconductors.

In recent years, Cross-Domain Recommendation (CDR) has drawn significant attention, which utilizes user data from multiple domains to enhance the recommendation performance. However, current CDR methods require sharing user data across domains, thereby violating the General Data Protection Regulation (GDPR). Consequently, numerous approaches have been proposed for Federated Cross-Domain Recommendation (FedCDR). Nevertheless, the data heterogeneity across different domains inevitably influences the overall performance of federated learning. In this study, we propose FedHCDR, a novel Federated Cross-Domain Recommendation framework with Hypergraph signal decoupling. Specifically, to address the data heterogeneity across domains, we introduce an approach called hypergraph signal decoupling (HSD) to decouple the user features into domain-exclusive and domain-shared features. The approach employs high-pass and low-pass hypergraph filters to decouple domain-exclusive and domain-shared user representations, which are trained by the local-global bi-directional transfer algorithm. In addition, a hypergraph contrastive learning (HCL) module is devised to enhance the learning of domain-shared user relationship information by perturbing the user hypergraph. Extensive experiments conducted on three real-world scenarios demonstrate that FedHCDR outperforms existing baselines significantly., Comment: 16 pages, 5 figures

Urban rail transit is a fundamental component of public transportation, however, commonly station-based path search algorithms often overlook the impact of transfer times on search results, leading to decreased accuracy. To solve this problem, this paper proposes a novel shortest path query algorithm based on adaptive topology optimization called the Adaptive Topology Extension Road Network Structure (ATEN). This algorithm categorizes transfer stations into different types and treats travel time and transfer time equivalently as weights for edges in the topological graph. The proposed algorithm introduces virtual stations to differentiate between pedestrian paths and train paths, eliminating the need for additional operations on transfer stations. The algorithm controls the extent of expansion in the urban rail transit topology, overcoming query errors caused by mishandling of transfer stations in the existing algorithm. Finally, a series of simulation experiments were conducted on Beijing's urban rail transit network to validate both correctness and efficiency of the proposed adaptive topology optimization algorithm. The results demonstrate significant advantages compared to existing similar algorithms.