Search

Graph Neural Networks (GNNs) with equivariant properties have achieved significant success in modeling complex dynamic systems and molecular properties. However, their expressiveness ability is limited by: (1) Existing methods often overlook the over-smoothing issue caused by traditional GNN models, as well as the gradient explosion or vanishing problems in deep GNNs. (2) Most models operate on first-order information, neglecting that the real world often consists of second-order systems, which further limits the model's representation capabilities. To address these issues, we propose the \textbf{Du}al \textbf{S}econd-order \textbf{E}quivariant \textbf{G}raph \textbf{O}rdinary Differential Equation (\method{}) for equivariant representation. Specifically, \method{} apply the dual second-order equivariant graph ordinary differential equations (Graph ODEs) on graph embeddings and node coordinates, simultaneously. Theoretically, we first prove that \method{} maintains the equivariant property. Furthermore, we provide theoretical insights showing that \method{} effectively alleviates the over-smoothing problem in both feature representation and coordinate update. Additionally, we demonstrate that the proposed \method{} mitigates the exploding and vanishing gradients problem, facilitating the training of deep multi-layer GNNs. Extensive experiments on benchmark datasets validate the superiority of the proposed \method{} compared to baselines.

Understanding the circumgalactic medium (CGM) distribution of galaxies is the key to revealing the dynamical exchange of materials between galaxies and their surroundings. In this work, we use DESI EDR dataset to investigate the cool CGM of galaxies ($0.3 10.0$, the EW(Mg II) at the minor axis is largely suppressed with respect to low mass galaxies. This suggests that the competing processes, such as stellar feedback and gravity, play a key role in shaping the distribution of outflowing gas., Comment: 20 pages, 17 figures. Submitted to ApJ. Comments are welcome

Early fault detection and timely maintenance scheduling can significantly mitigate operational risks in NPPs and enhance the reliability of operator decision-making. Therefore, it is necessary to develop an efficient Prognostics and Health Management (PHM) multi-step prediction model for predicting of system health status and prompt execution of maintenance operations. In this study, we propose a novel predictive model that integrates reinforcement learning with Long Short-Term Memory (LSTM) neural networks and the Expert Fuzzy Evaluation Method. The model is validated using parameter data for 20 different breach sizes in the Main Steam Line Break (MSLB) accident condition of the CPR1000 pressurized water reactor simulation model and it demonstrates a remarkable capability in accurately forecasting NPP parameter changes up to 128 steps ahead (with a time interval of 10 seconds per step, i.e., 1280 seconds), thereby satisfying the temporal advance requirement for fault prognostics in NPPs. Furthermore, this method provides an effective reference solution for PHM applications such as anomaly detection and remaining useful life prediction.

The MUltiplexed Survey Telescope (MUST) is a 6.5-meter telescope under development. Dedicated to highly-multiplexed, wide-field spectroscopic surveys, MUST observes over 20,000 targets simultaneously using 6.2-mm pitch positioning robots within a ~5 deg2 field of view. MUST aims to carry out the first Stage-V spectroscopic survey in the 2030s to map the 3D Universe with over 100 million galaxies and quasars, spanning from the nearby Universe to redshift z~5.5, corresponding to around 1 billion years after the Big Bang. To cover this extensive redshift range, we present an initial conceptual target selection algorithm for different types of galaxies, from local bright galaxies, luminous red galaxies, and emission line galaxies to high-redshift (2 < z < 5.5) Lyman-break galaxies. Using Fisher forecasts, we demonstrate that MUST can address fundamental questions in cosmology, including the nature of dark energy, test of gravity theories, and investigations into primordial physics. This is the first paper in the series of science white papers for MUST, with subsequent developments focusing on additional scientific cases such as galaxy and quasar evolution, Milky Way physics, and dynamic phenomena in the time-domain Universe., Comment: To be submitted to SCPMA

Utilizing fault diagnosis methods is crucial for nuclear power professionals to achieve efficient and accurate fault diagnosis for nuclear power plants (NPPs). The performance of traditional methods is limited by their dependence on complex feature extraction and skilled expert knowledge, which can be time-consuming and subjective. This paper proposes a novel intelligent fault diagnosis method for NPPs that combines enhanced temporal convolutional network (ETCN) with sparrow search algorithm (SSA). ETCN utilizes temporal convolutional network (TCN), self-attention (SA) mechanism and residual block for enhancing performance. ETCN excels at extracting local features and capturing time series information, while SSA adaptively optimizes its hyperparameters for superior performance. The proposed method's performance is experimentally verified on a CPR1000 simulation dataset. Compared to other advanced intelligent fault diagnosis methods, the proposed one demonstrates superior performance across all evaluation metrics. This makes it a promising tool for NPP intelligent fault diagnosis, ultimately enhancing operational reliability.

The safe and reliable operation of complex electromechanical systems in nuclear power plants is crucial for the safe production of nuclear power plants and their nuclear power unit. Therefore, accurate and timely fault diagnosis of nuclear power systems is of great significance for ensuring the safe and reliable operation of nuclear power plants. The existing fault diagnosis methods mainly target a single device or subsystem, making it difficult to analyze the inherent connections and mutual effects between different types of faults at the entire unit level. This article uses the AP1000 full-scale simulator to simulate the important mechanical component failures of some key systems in the primary and secondary circuits of nuclear power units, and constructs a fault dataset. Meanwhile, a hierarchical multi granularity classification fault diagnosis model based on the EfficientNet large model is proposed, aiming to achieve hierarchical classification of nuclear power faults. The results indicate that the proposed fault diagnosis model can effectively classify faults in different circuits and system components of nuclear power units into hierarchical categories. However, the fault dataset in this study was obtained from a simulator, which may introduce additional information due to parameter redundancy, thereby affecting the diagnostic performance of the model.

We study point impurities in non-relativistic quantum field theories, with a focus on scale-invariant fixed points. We establish the framework of conformal defects in Schr\"{o}dinger field theories and their correspondence to many-body states in a harmonic trap. We discuss a multi-critical quantization of the one-body wave function induced by an impurity, with examples of free and unitary Fermi gases. Our results can potentially be tested in experimental setups., Comment: 14 pages, 6 figures

The rapid development of generative Artificial Intelligence (AI) continually unveils the potential of Semantic Communication (SemCom). However, current talking-face SemCom systems still encounter challenges such as low bandwidth utilization, semantic ambiguity, and diminished Quality of Experience (QoE). This study introduces a Large Generative Model-assisted Talking-face Semantic Communication (LGM-TSC) System tailored for the talking-face video communication. Firstly, we introduce a Generative Semantic Extractor (GSE) at the transmitter based on the FunASR model to convert semantically sparse talking-face videos into texts with high information density. Secondly, we establish a private Knowledge Base (KB) based on the Large Language Model (LLM) for semantic disambiguation and correction, complemented by a joint knowledge base-semantic-channel coding scheme. Finally, at the receiver, we propose a Generative Semantic Reconstructor (GSR) that utilizes BERT-VITS2 and SadTalker models to transform text back into a high-QoE talking-face video matching the user's timbre. Simulation results demonstrate the feasibility and effectiveness of the proposed LGM-TSC system.

Meta-learning is a general approach to equip machine learning models with the ability to handle few-shot scenarios when dealing with many tasks. Most existing meta-learning methods work based on the assumption that all tasks are of equal importance. However, real-world applications often present heterogeneous tasks characterized by varying difficulty levels, noise in training samples, or being distinctively different from most other tasks. In this paper, we introduce a novel meta-learning method designed to effectively manage such heterogeneous tasks by employing rank-based task-level learning objectives, Heterogeneous Tasks Robust Meta-learning (HeTRoM). HeTRoM is proficient in handling heterogeneous tasks, and it prevents easy tasks from overwhelming the meta-learner. The approach allows for an efficient iterative optimization algorithm based on bi-level optimization, which is then improved by integrating statistical guidance. Our experimental results demonstrate that our method provides flexibility, enabling users to adapt to diverse task settings and enhancing the meta-learner's overall performance., Comment: 13 pages, 4 figures

Enabling Large Language Models (LLMs) to handle a wider range of complex tasks (e.g., coding, math) has drawn great attention from many researchers. As LLMs continue to evolve, merely increasing the number of model parameters yields diminishing performance improvements and heavy computational costs. Recently, OpenAI's o1 model has shown that inference strategies (i.e., Test-time Compute methods) can also significantly enhance the reasoning capabilities of LLMs. However, the mechanisms behind these methods are still unexplored. In our work, to investigate the reasoning patterns of o1, we compare o1 with existing Test-time Compute methods (BoN, Step-wise BoN, Agent Workflow, and Self-Refine) by using OpenAI's GPT-4o as a backbone on general reasoning benchmarks in three domains (i.e., math, coding, commonsense reasoning). Specifically, first, our experiments show that the o1 model has achieved the best performance on most datasets. Second, as for the methods of searching diverse responses (e.g., BoN), we find the reward models' capability and the search space both limit the upper boundary of these methods. Third, as for the methods that break the problem into many sub-problems, the Agent Workflow has achieved better performance than Step-wise BoN due to the domain-specific system prompt for planning better reasoning processes. Fourth, it is worth mentioning that we have summarized six reasoning patterns of o1, and provided a detailed analysis on several reasoning benchmarks.

For decades, the Bj{\o}ntegaard Delta (BD) has been the metric for evaluating codec Rate-Distortion (R-D) performance. Yet, in most studies, BD is determined using just 4-5 R-D data points, could this be sufficient? As codecs and quality metrics advance, does the conventional BD estimation still hold up? Crucially, are the performance improvements of new codecs and tools genuine, or merely artifacts of estimation flaws? This paper addresses these concerns by reevaluating BD estimation. We present a novel approach employing a parameterized deep neural network to model R-D curves with high precision across various metrics, accompanied by a comprehensive R-D dataset. This approach both assesses the reliability of BD calculations and serves as a precise BD estimator. Our findings advocate for the adoption of rigorous R-D sampling and reliability metrics in future compression research to ensure the validity and reliability of results.

Interleaved multimodal comprehension and generation, enabling models to produce and interpret both images and text in arbitrary sequences, have become a pivotal area in multimodal learning. Despite significant advancements, the evaluation of this capability remains insufficient. Existing benchmarks suffer from limitations in data scale, scope, and evaluation depth, while current evaluation metrics are often costly or biased, lacking in reliability for practical applications. To address these challenges, we introduce MMIE, a large-scale knowledge-intensive benchmark for evaluating interleaved multimodal comprehension and generation in Large Vision-Language Models (LVLMs). MMIE comprises 20K meticulously curated multimodal queries, spanning 3 categories, 12 fields, and 102 subfields, including mathematics, coding, physics, literature, health, and arts. It supports both interleaved inputs and outputs, offering a mix of multiple-choice and open-ended question formats to evaluate diverse competencies. Moreover, we propose a reliable automated evaluation metric, leveraging a scoring model fine-tuned with human-annotated data and systematic evaluation criteria, aimed at reducing bias and improving evaluation accuracy. Extensive experiments demonstrate the effectiveness of our benchmark and metrics in providing a comprehensive evaluation of interleaved LVLMs. Specifically, we evaluate eight LVLMs, revealing that even the best models show significant room for improvement, with most achieving only moderate results. We believe MMIE will drive further advancements in the development of interleaved LVLMs. We publicly release our benchmark and code in https://mmie-bench.github.io/.

Variation of Arctic sea ice has significant impacts on polar ecosystems, transporting routes, coastal communities, and global climate. Tracing the change of sea ice at a finer scale is paramount for both operational applications and scientific studies. Recent pan-Arctic sea ice forecasting methods that leverage advances in artificial intelligence has made promising progress over numerical models. However, forecasting sea ice at higher resolutions is still under-explored. To bridge the gap, we propose a two-staged deep learning framework, IceDiff, to forecast sea ice concentration at finer scales. IceDiff first leverages an independently trained vision transformer to generate coarse yet superior forecasting over previous methods at a regular 25km x 25km grid. This high-quality sea ice forecasting can be utilized as reliable guidance for the next stage. Subsequently, an unconditional diffusion model pre-trained on sea ice concentration maps is utilized for sampling down-scaled sea ice forecasting via a zero-shot guided sampling strategy and a patch-based method. For the first time, IceDiff demonstrates sea ice forecasting with the 6.25km x 6.25km resolution. IceDiff extends the boundary of existing sea ice forecasting models and more importantly, its capability to generate high-resolution sea ice concentration data is vital for pragmatic usages and research., Comment: 9 pages, 4 figures

Spiking Graph Networks (SGNs) have garnered significant attraction from both researchers and industry due to their ability to address energy consumption challenges in graph classification. However, SGNs are only effective for in-distribution data and cannot tackle out-of-distribution data. In this paper, we first propose the domain adaptation problem in SGNs, and introduce a novel framework named Degree-aware Spiking Graph Domain Adaptation for Classification. The proposed DeSGDA addresses the spiking graph domain adaptation problem by three aspects: node degree-aware personalized spiking representation, adversarial feature distribution alignment, and pseudo-label distillation. First, we introduce the personalized spiking representation method for generating degree-dependent spiking signals. Specifically, the threshold of triggering a spike is determined by the node degree, allowing this personalized approach to capture more expressive information for classification. Then, we propose the graph feature distribution alignment module that is adversarially trained using membrane potential against a domain discriminator. Such an alignment module can efficiently maintain high performance and low energy consumption in the case of inconsistent distribution. Additionally, we extract consistent predictions across two spaces to create reliable pseudo-labels, effectively leveraging unlabeled data to enhance graph classification performance. Extensive experiments on benchmark datasets validate the superiority of the proposed DeSGDA compared with competitive baselines.

Precipitation nowcasting plays a pivotal role in socioeconomic sectors, especially in severe convective weather warnings. Although notable progress has been achieved by approaches mining the spatiotemporal correlations with deep learning, these methods still suffer severe blurriness as the lead time increases, which hampers accurate predictions for extreme precipitation. To alleviate blurriness, researchers explore generative methods conditioned on blurry predictions. However, the pairs of blurry predictions and corresponding ground truth need to be generated in advance, making the training pipeline cumbersome and limiting the generality of generative models within blur modes that appear in training data. By rethinking the blurriness in precipitation nowcasting as a blur kernel acting on predictions, we propose an unsupervised postprocessing method to eliminate the blurriness without the requirement of training with the pairs of blurry predictions and corresponding ground truth. Specifically, we utilize blurry predictions to guide the generation process of a pre-trained unconditional denoising diffusion probabilistic model (DDPM) to obtain high-fidelity predictions with eliminated blurriness. A zero-shot blur kernel estimation mechanism and an auto-scale denoise guidance strategy are introduced to adapt the unconditional DDPM to any blurriness modes varying from datasets and lead times in precipitation nowcasting. Extensive experiments are conducted on 7 precipitation radar datasets, demonstrating the generality and superiority of our method.

Detecting deepfakes has become an important task. Most existing detection methods provide only real/fake predictions without offering human-comprehensible explanations. Recent studies leveraging MLLMs for deepfake detection have shown improvements in explainability. However, the performance of pre-trained MLLMs (e.g., LLaVA) remains limited due to a lack of understanding of their capabilities for this task and strategies to enhance them. In this work, we empirically assess the strengths and weaknesses of MLLMs specifically in deepfake detection via forgery features analysis. Building on these assessments, we propose a novel framework called ${X}^2$-DFD, consisting of three core modules. The first module, Model Feature Assessment (MFA), measures the detection capabilities of forgery features intrinsic to MLLMs, and gives a descending ranking of these features. The second module, Strong Feature Strengthening (SFS), enhances the detection and explanation capabilities by fine-tuning the MLLM on a dataset constructed based on the top-ranked features. The third module, Weak Feature Supplementing (WFS), improves the fine-tuned MLLM's capabilities on lower-ranked features by integrating external dedicated deepfake detectors. To verify the effectiveness of this framework, we further present a practical implementation, where an automated forgery features generation, evaluation, and ranking procedure is designed for MFA module; an automated generation procedure of the fine-tuning dataset containing real and fake images with explanations based on top-ranked features is developed for SFS model; an external conventional deepfake detector focusing on blending artifact, which corresponds to a low detection capability in the pre-trained MLLM, is integrated for WFS module. Experiments show that our approach enhances both detection and explanation performance.

In StyleGAN, convolution kernels are shaped by both static parameters shared across images and dynamic modulation factors $w^+\in\mathcal{W}^+$ specific to each image. Therefore, $\mathcal{W}^+$ space is often used for image inversion and editing. However, pre-trained model struggles with synthesizing out-of-domain images due to the limited capabilities of $\mathcal{W}^+$ and its resultant kernels, necessitating full fine-tuning or adaptation through a complex hypernetwork. This paper proposes an efficient refining strategy for dynamic kernels. The key idea is to modify kernels by low-rank residuals, learned from input image or domain guidance. These residuals are generated by matrix multiplication between two sets of tokens with the same number, which controls the complexity. We validate the refining scheme in image inversion and domain adaptation. In the former task, we design grouped transformer blocks to learn these token sets by one- or two-stage training. In the latter task, token sets are directly optimized to support synthesis in the target domain while preserving original content. Extensive experiments show that our method achieves low distortions for image inversion and high quality for out-of-domain editing., Comment: Accepted by PRCV2024

We present paper II comprising a 35 arcmin$^2$ JWST/NIRCam imaging and wide-field slitless spectroscopy mosaic centered on J0305$-$3150, a luminous quasar at $z=6.61$. The F356W grism data reveals 124 [OIII]+H$\beta$ emitters at $5.3

ASPIRE (A SPectroscopic survey of bIased halos in the Reionization Era) is a quasar legacy survey primarily using JWST to target a sample of 25 $z>6$ quasars with NIRCam slitless spectroscopy and imaging. The first study in this series found evidence of a strong overdensity of galaxies around J0305-3150, a luminous quasar at $z=6.61$, within a single NIRCam pointing obtained in JWST Cycle 1. Here, we present the first results of a JWST Cycle 2 mosaic that covers 35 arcmin$^2$ with NIRCam imaging/WFSS of the same field to investigate the spatial extent of the putative protocluster. The F356W grism data targets [OIII]+H$\beta$ at $5.310$ cMpc., Comment: submitted to ApJ june 2024. 18 pages, 7 figures. companion paper II by Champagne+24b is also available on arXiv today

Understanding when and how reionization happened is crucial for studying the early structure formation and the properties of first galaxies in the Universe. At $z>5.5$, the observed IGM optical depth shows a significant scatter, indicating an inhomogeneous reionization process. However, the nature of the inhomogeneous reionization remains debated. ASPIRE is a JWST Cycle 1 program that has spectroscopically identified $>400$ [OIII] emitters in 25 quasar fields at $z>6.5$. Combined with deep ground-based optical spectroscopy of ASPIRE quasars, ASPIRE program provides the current largest sample for IGM-galaxy connection studies during cosmic reionization. We present the first results of IGM effective optical depth measurements around [OIII] emitters using 14 ASPIRE quasar fields. We find the IGM transmission is tightly related with reionization-era galaxies to the extent that significant excess of Ly$\alpha$ transmission exists around [OIII] emitters. We measure the stacked IGM effective optical depth of IGM patches associated with [OIII] emitters and find they reach the same IGM effective optical depth at least dz~0.1 ahead of those IGM patches where no [OIII] emitters are detected, supporting earlier reionization around [OIII] emitters. Our results indicate an enhancement in IGM Ly$\alpha$ transmission around [OIII] emitters at scales beyond 25 $h^{-1}$ cMpc, consistent with the predicted topology of reionization from fluctuating UV background (UVB) models., Comment: 27 pages, 19 figures, 1 table, accepted for publication in ApJ

Diffusion-based generative models have demonstrated their powerful performance across various tasks, but this comes at a cost of the slow sampling speed. To achieve both efficient and high-quality synthesis, various distillation-based accelerated sampling methods have been developed recently. However, they generally require time-consuming fine tuning with elaborate designs to achieve satisfactory performance in a specific number of function evaluation (NFE), making them difficult to employ in practice. To address this issue, we propose Simple and Fast Distillation (SFD) of diffusion models, which simplifies the paradigm used in existing methods and largely shortens their fine-tuning time up to 1000$\times$. We begin with a vanilla distillation-based sampling method and boost its performance to state of the art by identifying and addressing several small yet vital factors affecting the synthesis efficiency and quality. Our method can also achieve sampling with variable NFEs using a single distilled model. Extensive experiments demonstrate that SFD strikes a good balance between the sample quality and fine-tuning costs in few-step image generation task. For example, SFD achieves 4.53 FID (NFE=2) on CIFAR-10 with only 0.64 hours of fine-tuning on a single NVIDIA A100 GPU. Our code is available at https://github.com/zju-pi/diff-sampler., Comment: Accepted by NeurIPS 2024

Conditional image synthesis based on user-specified requirements is a key component in creating complex visual content. In recent years, diffusion-based generative modeling has become a highly effective way for conditional image synthesis, leading to exponential growth in the literature. However, the complexity of diffusion-based modeling, the wide range of image synthesis tasks, and the diversity of conditioning mechanisms present significant challenges for researchers to keep up with rapid developments and understand the core concepts on this topic. In this survey, we categorize existing works based on how conditions are integrated into the two fundamental components of diffusion-based modeling, i.e., the denoising network and the sampling process. We specifically highlight the underlying principles, advantages, and potential challenges of various conditioning approaches in the training, re-purposing, and specialization stages to construct a desired denoising network. We also summarize six mainstream conditioning mechanisms in the essential sampling process. All discussions are centered around popular applications. Finally, we pinpoint some critical yet still open problems to be solved in the future and suggest some possible solutions. Our reviewed works are itemized at https://github.com/zju-pi/Awesome-Conditional-Diffusion-Models.

Large language models (LLMs) have exhibited remarkable capabilities across various domains and tasks, pushing the boundaries of our knowledge in learning and cognition. The latest model, OpenAI's o1, stands out as the first LLM with an internalized chain-of-thought technique using reinforcement learning strategies. While it has demonstrated surprisingly strong capabilities on various general language tasks, its performance in specialized fields such as medicine remains unknown. To this end, this report provides a comprehensive exploration of o1 on different medical scenarios, examining 3 key aspects: understanding, reasoning, and multilinguality. Specifically, our evaluation encompasses 6 tasks using data from 37 medical datasets, including two newly constructed and more challenging question-answering (QA) tasks based on professional medical quizzes from the New England Journal of Medicine (NEJM) and The Lancet. These datasets offer greater clinical relevance compared to standard medical QA benchmarks such as MedQA, translating more effectively into real-world clinical utility. Our analysis of o1 suggests that the enhanced reasoning ability of LLMs may (significantly) benefit their capability to understand various medical instructions and reason through complex clinical scenarios. Notably, o1 surpasses the previous GPT-4 in accuracy by an average of 6.2% and 6.6% across 19 datasets and two newly created complex QA scenarios. But meanwhile, we identify several weaknesses in both the model capability and the existing evaluation protocols, including hallucination, inconsistent multilingual ability, and discrepant metrics for evaluation. We release our raw data and model outputs at https://ucsc-vlaa.github.io/o1_medicine/ for future research., Comment: The first four authors contributed equally, project page available at https://ucsc-vlaa.github.io/o1_medicine/

Recent advancements in multimodal large language models (MLLMs) have aimed to integrate and interpret data across diverse modalities. However, the capacity of these models to concurrently process and reason about multiple modalities remains inadequately explored, partly due to the lack of comprehensive modality-wise benchmarks. We introduce OmniBench, a novel benchmark designed to rigorously evaluate models' ability to recognize, interpret, and reason across visual, acoustic, and textual inputs simultaneously. We define models capable of such tri-modal processing as omni-language models (OLMs). OmniBench is distinguished by high-quality human annotations, ensuring that accurate responses require integrated understanding and reasoning across all three modalities. Our main findings reveal that: i) most OLMs exhibit critical limitations in instruction-following and reasoning capabilities within tri-modal contexts; and ii) most baselines models perform poorly (below 50\% accuracy) even when provided with alternative textual representations of images or/and audio. These results suggest that the ability to construct a consistent context from text, image, and audio is often overlooked in existing MLLM training paradigms. To address this gap, we curate an instruction tuning dataset of 84.5K training samples, OmniInstruct, for training OLMs to adapt to multimodal contexts. We advocate for future research to focus on developing more robust tri-modal integration techniques and training strategies to enhance OLM performance across diverse modalities. The codes and live leaderboard could be found at https://m-a-p.ai/OmniBench.

Neural networks (NN) are extensively studied in cutting-edge soft sensor models due to their feature extraction and function approximation capabilities. Current research into network-based methods primarily focuses on models' offline accuracy. Notably, in industrial soft sensor context, online optimizing stability and interpretability are prioritized, followed by accuracy. This requires a clearer understanding of network's training process. To bridge this gap, we propose a novel NN named the Approximated Orthogonal Projection Unit (AOPU) which has solid mathematical basis and presents superior training stability. AOPU truncates the gradient backpropagation at dual parameters, optimizes the trackable parameters updates, and enhances the robustness of training. We further prove that AOPU attains minimum variance estimation (MVE) in NN, wherein the truncated gradient approximates the natural gradient (NG). Empirical results on two chemical process datasets clearly show that AOPU outperforms other models in achieving stable convergence, marking a significant advancement in soft sensor field.

The quantity of processed data is crucial for advancing the field of singing voice synthesis. While there are tools available for lyric or note transcription tasks, they all need pre-processed data which is relatively time-consuming (e.g., vocal and accompaniment separation). Besides, most of these tools are designed to address a single task and struggle with aligning lyrics and notes (i.e., identifying the corresponding notes of each word in lyrics). To address those challenges, we first design a pipeline by optimizing existing tools and annotating numerous lyric-note pairs of songs. Then, based on the annotated data, we train a unified SongTrans model that can directly transcribe lyrics and notes while aligning them simultaneously, without requiring pre-processing songs. Our SongTrans model consists of two modules: (1) the \textbf{Autoregressive module} predicts the lyrics, along with the duration and note number corresponding to each word in a lyric. (2) the \textbf{Non-autoregressive module} predicts the pitch and duration of the notes. Our experiments demonstrate that SongTrans achieves state-of-the-art (SOTA) results in both lyric and note transcription tasks. Furthermore, it is the first model capable of aligning lyrics with notes. Experimental results demonstrate that the SongTrans model can effectively adapt to different types of songs (e.g., songs with accompaniment), showcasing its versatility for real-world applications.

The topological Hall effect (THE) has attracted great attention since it provides an important probe of the interaction between electron and topological spin textures. THE has been considered an experimental signature of the topological spin texture of skyrmions. While THE has been widely reported in chiral magnets, oxide heterostructures, and hybrid systems such as ferromagnet/heavy metal and ferromagnet/topological insulators, the study of monolayer structures is lacking, hindering the understanding of noncollinear spin textures at the atomically thin scale. Here, we show a discernible THE via proximity coupling of Fe-doped monolayer WSe2 (Fe:WSe2) synthesized using chemical vapor deposition on a Pt Hall bar. Multiple characterization methods were employed to demonstrate that Fe atoms substitutionally replace W atoms, making a two-dimensional (2D) van der Waals (vdW) dilute magnetic semiconductor (DMS) at room temperature. Distinct from the intrinsic anomalous Hall effect, we found the transverse Hall resistivity of Fe:WSe2 displaying two additional dip/peak features in the temperature-dependent measurements, consistent with the contribution of THE. The topological Hall effect is attributed to the magnetic skyrmions that emerge from the Dzyaloshinskii-Moriya interactions at the Fe:WSe2 and Pt interface. Our work shows that a DMS synthesized from 2D vdW transition metal dichalcogenides is promising for realizing magnetic skyrmions and spintronic applications.

3T TRAM with doping profile (P+PNPNN+) is experimentally demonstrated on a silicon photonic platform. By using additional implant layers, this device provides excellent memory performance compared to the conventional structure (PNPN). TCAD is used to reflect the physical behavior, and the high-speed memory operations are described through the model., Comment: 4 pages, 15 figures

Multimodal Large Language Models (MLLMs) are evaluated on various benchmarks, such as image captioning, visual question answering, and reasoning. However, many of these benchmarks include overly simple or uninformative samples, complicating the effective distinction of different MLLMs' performance. Furthermore, evaluating models across numerous benchmarks incurs a significant computational burden. To address these issues, we propose LIME (Less Is More for MLLM Evaluation), a refined and efficient benchmark curated through a semi-automated pipeline. This pipeline filters out uninformative samples and eliminates answer leakage by focusing on tasks that necessitate image-based understanding. Our experiments indicate that LIME reduces the number of samples by 76% and evaluation time by 77%, while also providing a more effective means of distinguishing the capabilities of different models. Notably, we find that traditional automatic metrics, such as CIDEr, are inadequate for assessing MLLMs' captioning performance; excluding the caption task score yields a more accurate reflection of overall model performance. All code and data are available at https://github.com/kangreen0210/LIME.

Explainable fake news detection predicts the authenticity of news items with annotated explanations. Today, Large Language Models (LLMs) are known for their powerful natural language understanding and explanation generation abilities. However, presenting LLMs for explainable fake news detection remains two main challenges. Firstly, fake news appears reasonable and could easily mislead LLMs, leaving them unable to understand the complex news-faking process. Secondly, utilizing LLMs for this task would generate both correct and incorrect explanations, which necessitates abundant labor in the loop. In this paper, we propose LLM-GAN, a novel framework that utilizes prompting mechanisms to enable an LLM to become Generator and Detector and for realistic fake news generation and detection. Our results demonstrate LLM-GAN's effectiveness in both prediction performance and explanation quality. We further showcase the integration of LLM-GAN to a cloud-native AI platform to provide better fake news detection service in the cloud.

In recent years, the multimedia forensics and security community has seen remarkable progress in multitask learning for DeepFake (i.e., face forgery) detection. The prevailing strategy has been to frame DeepFake detection as a binary classification problem augmented by manipulation-oriented auxiliary tasks. This strategy focuses on learning features specific to face manipulations, which exhibit limited generalizability. In this paper, we delve deeper into semantics-oriented multitask learning for DeepFake detection, leveraging the relationships among face semantics via joint embedding. We first propose an automatic dataset expansion technique that broadens current face forgery datasets to support semantics-oriented DeepFake detection tasks at both the global face attribute and local face region levels. Furthermore, we resort to joint embedding of face images and their corresponding labels (depicted by textual descriptions) for prediction. This approach eliminates the need for manually setting task-agnostic and task-specific parameters typically required when predicting labels directly from images. In addition, we employ a bi-level optimization strategy to dynamically balance the fidelity loss weightings of various tasks, making the training process fully automated. Extensive experiments on six DeepFake datasets show that our method improves the generalizability of DeepFake detection and, meanwhile, renders some degree of model interpretation by providing human-understandable explanations.

Based on our previous study [S. Wang $\textit{et al}$. J. Chem. Phys. $\textbf{153}$, 184102 (2020)], we generalize the theory of molecular emission power spectra (EPS) from one molecule to multichromophoric systems in the framework of macroscopic quantum electrodynamics. This generalized theory is applicable to ensembles of molecules, providing a comprehensive description of the molecular spontaneous emission spectrum in arbitrary inhomogeneous, dispersive, and absorbing media. In the far-field region, the analytical formula of EPS can be expressed as the product of a lineshape function (LF) and an electromagnetic environment factor (EEF). To demonstrate the polaritonic effect on multichromophoric systems, we simulate the LF and EEF for one to three molecules weakly coupled to surface plasmon polaritons above a silver surface. Our analytical expressions show that the peak broadening originates from not only the spontaneous emission rates but also the imaginary part of resonant dipole-dipole interactions (non-Hermitian interactions), which is associated with the superradiance of molecular aggregates, indicating that the superradiance rate can be controlled through an intermolecular distance and the design of dielectric environments. This study presents an alternative approach to directly analyze the hybrid-state dynamics of multichromophoric systems coupled with polaritons., Comment: 19 pages, 5 figures, submitted to The Journal of Chemical Physics

Diffusion models have been widely utilized for image restoration. However, previous blind image restoration methods still need to assume the type of degradation model while leaving the parameters to be optimized, limiting their real-world applications. Therefore, we aim to tame generative diffusion prior for universal blind image restoration dubbed BIR-D, which utilizes an optimizable convolutional kernel to simulate the degradation model and dynamically update the parameters of the kernel in the diffusion steps, enabling it to achieve blind image restoration results even in various complex situations. Besides, based on mathematical reasoning, we have provided an empirical formula for the chosen of adaptive guidance scale, eliminating the need for a grid search for the optimal parameter. Experimentally, Our BIR-D has demonstrated superior practicality and versatility than off-the-shelf unsupervised methods across various tasks both on real-world and synthetic datasets, qualitatively and quantitatively. BIR-D is able to fulfill multi-guidance blind image restoration. Moreover, BIR-D can also restore images that undergo multiple and complicated degradations, demonstrating the practical applications., Comment: 14 pages, 9 figures, 8 tables

Recent research on knowledge distillation has increasingly focused on logit distillation because of its simplicity, effectiveness, and versatility in model compression. In this paper, we introduce Refined Logit Distillation (RLD) to address the limitations of current logit distillation methods. Our approach is motivated by the observation that even high-performing teacher models can make incorrect predictions, creating a conflict between the standard distillation loss and the cross-entropy loss. This conflict can undermine the consistency of the student model's learning objectives. Previous attempts to use labels to empirically correct teacher predictions may undermine the class correlation. In contrast, our RLD employs labeling information to dynamically refine teacher logits. In this way, our method can effectively eliminate misleading information from the teacher while preserving crucial class correlations, thus enhancing the value and efficiency of distilled knowledge. Experimental results on CIFAR-100 and ImageNet demonstrate its superiority over existing methods. The code is provided at \text{https://github.com/zju-SWJ/RLD}., Comment: 11 pages, 7 figures

This paper presents the results of the shared task on Chinese metaphor generation, hosted at the 13th CCF Conference on Natural Language Processing and Chinese Computing (NLPCC 2024). The goal of this shared task is to generate Chinese metaphors using machine learning techniques and effectively identifying basic components of metaphorical sentences. It is divided into two subtasks: 1) Metaphor Generation, which involves creating a metaphor from a provided tuple consisting of TENOR, GROUND, and VEHICLE. The goal here is to synthesize a metaphor that connects the subject (i.e. TENOR) with the object (i.e. VEHICLE), guided by the concept of the GROUND. 2) Metaphor Components Identification, which extracts the most fitting TENORs, GROUNDs, and VEHICLEs from a metaphorical sentence. This component requires the identification of the most fitting metaphor elements that correspond to the specified grounds. In addition to overall results, we report on the setup and insights from the metaphor generation shared task, which attracted a total of 4 participating teams across both subtasks.

The CNN has achieved excellent results in the automatic classification of medical images. In this study, we propose a novel deep residual 3D attention non-local network (NL-RAN) to classify CT images included COVID-19, common pneumonia, and normal to perform rapid and explainable COVID-19 diagnosis. We built a deep residual 3D attention non-local network that could achieve end-to-end training. The network is embedded with a nonlocal module to capture global information, while a 3D attention module is embedded to focus on the details of the lesion so that it can directly analyze the 3D lung CT and output the classification results. The output of the attention module can be used as a heat map to increase the interpretability of the model. 4079 3D CT scans were included in this study. Each scan had a unique label (novel coronavirus pneumonia, common pneumonia, and normal). The CT scans cohort was randomly split into a training set of 3263 scans, a validation set of 408 scans, and a testing set of 408 scans. And compare with existing mainstream classification methods, such as CovNet, CBAM, ResNet, etc. Simultaneously compare the visualization results with visualization methods such as CAM. Model performance was evaluated using the Area Under the ROC Curve(AUC), precision, and F1-score. The NL-RAN achieved the AUC of 0.9903, the precision of 0.9473, and the F1-score of 0.9462, surpass all the classification methods compared. The heat map output by the attention module is also clearer than the heat map output by CAM. Our experimental results indicate that our proposed method performs significantly better than existing methods. In addition, the first attention module outputs a heat map containing detailed outline information to increase the interpretability of the model. Our experiments indicate that the inference of our model is fast. It can provide real-time assistance with diagnosis.

Image inpainting, which is the task of filling in missing areas in an image, is a common image editing technique. Inpainting can be used to conceal or alter image contents in malicious manipulation of images, driving the need for research in image inpainting detection. Existing methods mostly rely on a basic encoder-decoder structure, which often results in a high number of false positives or misses the inpainted regions, especially when dealing with targets of varying semantics and scales. Additionally, the absence of an effective approach to capture boundary artifacts leads to less accurate edge localization. In this paper, we describe a new method for inpainting detection based on a Dense Feature Interaction Network (DeFI-Net). DeFI-Net uses a novel feature pyramid architecture to capture and amplify multi-scale representations across various stages, thereby improving the detection of image inpainting by better revealing feature-level interactions. Additionally, the network can adaptively direct the lower-level features, which carry edge and shape information, to refine the localization of manipulated regions while integrating the higher-level semantic features. Using DeFI-Net, we develop a method combining complementary representations to accurately identify inpainted areas. Evaluation on five image inpainting datasets demonstrate the effectiveness of our approach, which achieves state-of-the-art performance in detecting inpainting across diverse models.

In this paper, we propose an improved method based on Mixed-Integer Linear Programming/Mixed-Integer Quadratic Constraint Programming (MILP/MIQCP) to automatically find better differential-linear (DL) distinguishers for the all members of Simon and Simeck block cipher families. To be specific, we first give the completely precise MILP model to describe the linear part, and explain how to utilize the general expressions of \textsf{Gurobi} solver to model the propagation of continuous difference for the middle part in a quite easy way. Secondly, in order to solve the MILP/MIQCP model in a reasonable time, we propose two heuristic strategies based on the divide-and-conquer idea to speed up the search process. Thirdly, we introduce the transforming technique, which exploits the clustering effect on DL trails, to improve the estimated correlation of the DL approximation. We apply our method to Simon and Simeck block cipher families. Consequently, we find the 14/17/21/26-round theoretical DL distinguishers of Simon32/48/64/96, which extend the previous longest ones of Simon32/48/96 by one round and Simon64 by two rounds, respectively. For Simeck, we do not explore longer distinguishers compared to the currently best results, but refresh all the results of Zhou et al. (the first work to automate finding DL distinguishers for Simon-like ciphers using MILP/MIQCP). Besides, in order to validate the correctness of these distinguishers, the experimental verifications are conducted on Simon32/Simeck32 and Simon48/Simeck48. The results show that our theoretical estimations on correlations are very close to the experimental ones, which can be regarded as a concrete support for the effectiveness of our method., Comment: 37 pages

We present the identification of changing-look active galactic nuclei (CL-AGNs) from the Dark Energy Spectroscopic Instrument First Data Release and Sloan Digital Sky Survey Data Release 16 at z \leq 0.9. To confirm the CL-AGNs, we utilize spectral flux calibration assessment via an [O\,{\sc iii}]-based calibration, pseudo-photometry examination, and visual inspection. This rigorous selection process allows us to compile a statistical catalog of 561 CL-AGNs, encompassing 527 $\rm H\beta$, 149$\rm H\alpha$, and 129 Mg II CL behaviors. In this sample, we find 1) a 283:278 ratio of turn-on to turn-off CL-AGNs. 2) the critical value for CL events is confirmed around Eddington ratio \sim 0.01. 3) a strong correlation between the change in the luminosity of the broad emission lines (BEL) and variation in the continuum luminosity, with Mg II and $\rm H\beta$ displaying similar responses during CL phases. 4) the Baldwin-Phillips-Terlevich diagram for CL-AGNs shows no statistically difference from the general AGN catalog. 5) five CL-AGNs are associated with asymmetrical mid-infrared flares, possibly linked to tidal disruption events. Given the large CL-AGNs and the stochastic sampling of spectra, we propose that some CL events are inherently due to typical AGN variability during low accretion rates, particularly for CL events of the singular BEL. Finally, we introduce a Peculiar CL phase, characterized by a gradual decline over decades in the light curve and the complete disappearance of entire BEL in faint spectra, indicative of a real transition in the accretion disk., Comment: Submitted to ApJS, comments welcome

Handwriting Verification is a critical in document forensics. Deep learning based approaches often face skepticism from forensic document examiners due to their lack of explainability and reliance on extensive training data and handcrafted features. This paper explores using Vision Language Models (VLMs), such as OpenAI's GPT-4o and Google's PaliGemma, to address these challenges. By leveraging their Visual Question Answering capabilities and 0-shot Chain-of-Thought (CoT) reasoning, our goal is to provide clear, human-understandable explanations for model decisions. Our experiments on the CEDAR handwriting dataset demonstrate that VLMs offer enhanced interpretability, reduce the need for large training datasets, and adapt better to diverse handwriting styles. However, results show that the CNN-based ResNet-18 architecture outperforms the 0-shot CoT prompt engineering approach with GPT-4o (Accuracy: 70%) and supervised fine-tuned PaliGemma (Accuracy: 71%), achieving an accuracy of 84% on the CEDAR AND dataset. These findings highlight the potential of VLMs in generating human-interpretable decisions while underscoring the need for further advancements to match the performance of specialized deep learning models., Comment: 4 Pages, 1 Figure, 1 Table, Accepted as Short paper at Irish Machine Vision and Image Processing (IMVIP) Conference

Weather forecasting refers to learning evolutionary patterns of some key upper-air and surface variables which is of great significance. Recently, deep learning-based methods have been increasingly applied in the field of weather forecasting due to their powerful feature learning capabilities. However, prediction methods based on the original space iteration struggle to effectively and efficiently utilize large number of weather variables. Therefore, we propose an 'encoding-prediction-decoding' prediction network. This network can efficiently benefit to more related input variables with key variables, that is, it can adaptively extract key variable-related low-dimensional latent feature from much more input atmospheric variables for iterative prediction. And we construct a loss function to guide the iteration of latent feature by utilizing multiple atmospheric variables in corresponding lead times. The obtained latent features through iterative prediction are then decoded to obtain the predicted values of key variables in multiple lead times. In addition, we improve the HTA algorithm in \cite{bi2023accurate} by inputting more time steps to enhance the temporal correlation between the prediction results and input variables. Both qualitative and quantitative prediction results on ERA5 dataset validate the superiority of our method over other methods. (The code will be available at https://github.com/rs-lsl/Kvp-lsi)

Discrete-Time Dynamic Graphs (DTDGs), which are prevalent in real-world implementations and notable for their ease of data acquisition, have garnered considerable attention from both academic researchers and industry practitioners. The representation learning of DTDGs has been extensively applied to model the dynamics of temporally changing entities and their evolving connections. Currently, DTDG representation learning predominantly relies on GNN+RNN architectures, which manifest the inherent limitations of both Graph Neural Networks (GNNs) and Recurrent Neural Networks (RNNs). GNNs suffer from the over-smoothing issue as the models architecture goes deeper, while RNNs struggle to capture long-term dependencies effectively. GNN+RNN architectures also grapple with scaling to large graph sizes and long sequences. Additionally, these methods often compute node representations separately and focus solely on individual node characteristics, thereby overlooking the behavior intersections between the two nodes whose link is being predicted, such as instances where the two nodes appear together in the same context or share common neighbors. This paper introduces a novel representation learning method DTFormer for DTDGs, pivoting from the traditional GNN+RNN framework to a Transformer-based architecture. Our approach exploits the attention mechanism to concurrently process topological information within the graph at each timestamp and temporal dynamics of graphs along the timestamps, circumventing the aforementioned fundamental weakness of both GNNs and RNNs. Moreover, we enhance the model's expressive capability by incorporating the intersection relationships among nodes and integrating a multi-patching module. Extensive experiments conducted on six public dynamic graph benchmark datasets confirm our model's efficacy, achieving the SOTA performance., Comment: 11 pages, 3 figures

Given the remarkable success that large visual language models (LVLMs) have achieved in image perception tasks, the endeavor to make LVLMs perceive the world like humans is drawing increasing attention. Current multi-modal benchmarks primarily focus on facts or specific topic-related knowledge contained within individual images. However, they often overlook the associative relations between multiple images, which require the identification and analysis of similarities among entities or content present in different images. Therefore, we propose the multi-image relation association task and a meticulously curated Multi-granularity Multi-image Relational Association (MMRA) benchmark, comprising 1,024 samples. In order to systematically and comprehensively evaluate current LVLMs, we establish an associational relation system among images that contain 11 subtasks (e.g, UsageSimilarity, SubEvent) at two granularity levels (i.e., image and entity) according to the relations in ConceptNet. Our experiments reveal that on the MMRA benchmark, current multi-image LVLMs exhibit distinct advantages and disadvantages across various subtasks. Notably, fine-grained, entity-level multi-image perception tasks pose a greater challenge for LVLMs compared to image-level tasks. Moreover, LVLMs perform poorly on spatial-related tasks, indicating that LVLMs still have limited spatial awareness. Additionally, our findings indicate that while LVLMs demonstrate a strong capability to perceive image details, enhancing their ability to associate information across multiple images hinges on improving the reasoning capabilities of their language model component. Moreover, we explored the ability of LVLMs to perceive image sequences within the context of our multi-image association task. Our experiments show that the majority of current LVLMs do not adequately model image sequences during the pre-training process., Comment: VLMs, Multi-Image Association

To overcome the imbalanced multimodal learning problem, where models prefer the training of specific modalities, existing methods propose to control the training of uni-modal encoders from different perspectives, taking the inter-modal performance discrepancy as the basis. However, the intrinsic limitation of modality capacity is ignored. The scarcely informative modalities can be recognized as ``worse-learnt'' ones, which could force the model to memorize more noise, counterproductively affecting the multimodal model ability. Moreover, the current modality modulation methods narrowly concentrate on selected worse-learnt modalities, even suppressing the training of others. Hence, it is essential to consider the intrinsic limitation of modality capacity and take all modalities into account during balancing. To this end, we propose the Diagnosing \& Re-learning method. The learning state of each modality is firstly estimated based on the separability of its uni-modal representation space, and then used to softly re-initialize the corresponding uni-modal encoder. In this way, the over-emphasizing of scarcely informative modalities is avoided. In addition, encoders of worse-learnt modalities are enhanced, simultaneously avoiding the over-training of other modalities. Accordingly, multimodal learning is effectively balanced and enhanced. Experiments covering multiple types of modalities and multimodal frameworks demonstrate the superior performance of our simple-yet-effective method for balanced multimodal learning. The source code and dataset are available at \url{https://github.com/GeWu-Lab/Diagnosing_Relearning_ECCV2024}., Comment: Accepted by ECCV 2024

Spatial transformations of light are ubiquitous in optics, with examples ranging from simple imaging with a lens to quantum and classical information processing in waveguide meshes. Multi-plane light converter (MPLC) systems have emerged as a platform that promises completely general spatial transformations, i.e., a universal unitary. However until now, MPLC systems have demonstrated transformations that are far from general, e.g., converting from a Gaussian to Laguerre-Gauss mode. Here, we demonstrate the promise of an MLPC, the ability to impose an arbitrary unitary transformation that can be reconfigured dynamically. Specifically, we consider transformations on superpositions of parallel free-space beams arranged in an array, which is a common information encoding in photonics. We experimentally test the full gamut of unitary transformations for a system of two parallel beams and make a map of their fidelity. We obtain an average transformation fidelity of $0.85 \pm 0.03$. This high-fidelity suggests MPLCs are a useful tool implementing the unitary transformations that comprise quantum and classical information processing.

A combined experimental and theoretical study is carried out on the three-body recombination process in a gas of microwave-shielded polar molecules. For ground-state polar molecules dressed with a strong microwave field, field-linked bound states can appear in the intermolecular potential. We model three-body recombination into such bound states using classical trajectory calculations. Our results show that recombination can explain the enhanced loss rates observed at small microwave detunings in trapped samples of bosonic NaCs [Bigagli, $\textit{et al.}$, Nat. Phys. $\textbf{19}$ 1579-1584 (2023)]. Specifically, our calculations reproduce the experimentally measured three-body loss rates across a wide range of microwave Rabi couplings, detunings, and temperatures. This work suggests that for bosonic shielded molecular systems in which the two-body loss is sufficiently suppressed and a field-linked bound state is present, the dominant loss process will be three-body recombination., Comment: 11 pages, 5 figures

Efficient fine-tuning plays a fundamental role in modern large models, with low-rank adaptation emerging as a particularly promising approach. However, the existing variants of LoRA are hampered by limited expressiveness, a tendency to overfit, and sensitivity to hyperparameter settings. This paper presents LoRA Slow Cascade Learning (LoRASC), an innovative technique designed to enhance LoRA's expressiveness and generalization capabilities while preserving its training efficiency. Our approach augments expressiveness through a cascaded learning strategy that enables a mixture-of-low-rank adaptation, thereby increasing the model's ability to capture complex patterns. Additionally, we introduce a slow-fast update mechanism and cascading noisy tuning to bolster generalization. The extensive experiments on various language and vision datasets, as well as robustness benchmarks, demonstrate that the proposed method not only significantly outperforms existing baselines, but also mitigates overfitting, enhances model stability, and improves OOD robustness. Code will be release in https://github.com/microsoft/LoRASC very soon.