arXiv subjects

Bing Wang

Publications and source records attributed to Bing Wang.

At least 127 records · Page 7Linked to original sources

arXiv2025

Non-r-partite graphs without complete split subgraphs

The classical Simonovits' chromatic critical edge theorem shows that for sufficiently large $n$, if $H$ is an edge-color-critical graph with $χ(H)=p+1\ge 3$, then the Turán graph $T_{n,p}$ is the unique extremal graph with respect to ${\rm ex}(n,H)$. Denote by ${\rm EX}_{r+1}(n,H)$ and ${\rm SPEX}_{r+1}(n,H)$ the family of $n$-vertex $H$-free non-$r$-partite graphs with the maximum size and with the spectral radius, respectively. Li and Peng [SIAM J. Discrete Math. 37 (2023) 2462--2485] characterized the unique graph in $\mathrm{SPEX}_{r+1}(n,K_{r+1})$ for $r\geq 2$ and showed that $\mathrm{SPEX}_{r+1}(n,K_{r+1})\subseteq \mathrm{EX}_{r+1}(n,K_{r+1})$. It is interesting to study the extremal or spectral extremal problems for color-critical graph $H$ in non-$r$-partite graphs. For $p\geq 2$ and $q\geq 1$, we call the graph $B_{p,q}:=K_p\nabla qK_1$ a complete split graph (or generalized book graph). In this note, we determine the unique spectral extremal graph in $\mathrm{SPEX}_{r+1}(n,B_{p,q})$ and show that $\mathrm{SPEX}_{r+1}(n,B_{p,q})\subseteq \mathrm{EX}_{r+1}(n,B_{p,q})$ for sufficiently large $n$.

math.CO↗

arXiv2025

Scaling Transformers for Discriminative Recommendation via Generative Pretraining

Discriminative recommendation tasks, such as CTR (click-through rate) and CVR (conversion rate) prediction, play critical roles in the ranking stage of large-scale industrial recommender systems. However, training a discriminative model encounters a significant overfitting issue induced by data sparsity. Moreover, this overfitting issue worsens with larger models, causing them to underperform smaller ones. To address the overfitting issue and enhance model scalability, we propose a framework named GPSD (\textbf{G}enerative \textbf{P}retraining for \textbf{S}calable \textbf{D}iscriminative Recommendation), drawing inspiration from generative training, which exhibits no evident signs of overfitting. GPSD leverages the parameters learned from a pretrained generative model to initialize a discriminative model, and subsequently applies a sparse parameter freezing strategy. Extensive experiments conducted on both industrial-scale and publicly available datasets demonstrate the superior performance of GPSD. Moreover, it delivers remarkable improvements in online A/B tests. GPSD offers two primary advantages: 1) it substantially narrows the generalization gap in model training, resulting in better test performance; and 2) it leverages the scalability of Transformers, delivering consistent performance gains as models are scaled up. Specifically, we observe consistent performance improvements as the model dense parameters scale from 13K to 0.3B, closely adhering to power laws. These findings pave the way for unifying the architectures of recommendation models and language models, enabling the direct application of techniques well-established in large language models to recommendation models. The code is available at https://github.com/chqiwang/gpsd-rec.

cs.IR↗

arXiv2025

Variety Is the Spice of Life: Detecting Misinformation with Dynamic Environmental Representations

The proliferation of misinformation across diverse social media platforms has drawn significant attention from both academic and industrial communities due to its detrimental effects. Accordingly, automatically distinguishing misinformation, dubbed as Misinformation Detection (MD), has become an increasingly active research topic. The mainstream methods formulate MD as a static learning paradigm, which learns the mapping between the content, links, and propagation of news articles and the corresponding manual veracity labels. However, the static assumption is often violated, since in real-world scenarios, the veracity of news articles may vacillate within the dynamically evolving social environment. To tackle this problem, we propose a novel framework, namely Misinformation detection with Dynamic Environmental Representations (MISDER). The basic idea of MISDER lies in learning a social environmental representation for each period and employing a temporal model to predict the representation for future periods. In this work, we specify the temporal model as the LSTM model, continuous dynamics equation, and pre-trained dynamics system, suggesting three variants of MISDER, namely MISDER-LSTM, MISDER-ODE, and MISDER-PT, respectively. To evaluate the performance of MISDER, we compare it to various MD baselines across 2 prevalent datasets, and the experimental results can indicate the effectiveness of our proposed model.

cs.CL↗

arXiv2025

A Flare-related Decimetric Type-IV Radio Burst Induced by the X2 Radiation of Electron Cyclotron Maser Emission

The radiation mechanism of decimetric wideband and pulsating radio bursts from the Sun (in terms of decimetric type-IV (t-IVdm) burst) and other flaring stars is a long-standing problem. Early investigations were based on the leading-spot hypothesis for the sun and yielded contradictory results. Here, we analyzed the flare-associated t-IVdm burst on 20110924 with medium-strong levels of polarization and from sources near a sunspot. We found that the emission is intermittent and the maximum $T_B$ exceeds 10$^{11}$ K, with well-defined upper and lower frequency cutoffs. The radio sources are left-handed polarized, located above the sunspot with a negative polarity. The sources align well with the sites of the second harmonic of the local electron gyrofrequency. These findings provide essential evidence that the burst is induced by the electron cyclotron maser emission (ECME) in the harmonic X mode. We further modeled the transport of downward-streaming energetic electrons along a coronal loop and found most electrons get mirrored within the specific altitude range of 20-100 Mm. This explains why such bursts tend to have well-defined spectral ranges. We also found the ECME-radiating energetic electrons exhibit a shell-like VDF instead of the generally-presumed loss-cone distribution. The study greatly expands the application of ECME in solar radio astronomy and provides solar samples for similar bursts from other flaring stars.

astro-ph.SR↗

arXiv2025

MMOne: Representing Multiple Modalities in One Scene

Humans perceive the world through multimodal cues to understand and interact with the environment. Learning a scene representation for multiple modalities enhances comprehension of the physical world. However, modality conflicts, arising from inherent distinctions among different modalities, present two critical challenges: property disparity and granularity disparity. To address these challenges, we propose a general framework, MMOne, to represent multiple modalities in one scene, which can be readily extended to additional modalities. Specifically, a modality modeling module with a novel modality indicator is proposed to capture the unique properties of each modality. Additionally, we design a multimodal decomposition mechanism to separate multi-modal Gaussians into single-modal Gaussians based on modality differences. We address the essential distinctions among modalities by disentangling multimodal information into shared and modality-specific components, resulting in a more compact and efficient multimodal scene representation. Extensive experiments demonstrate that our method consistently enhances the representation capability for each modality and is scalable to additional modalities. The code is available at https://github.com/Neal2020GitHub/MMOne.

cs.CV↗

arXiv2025

DriveMRP: Enhancing Vision-Language Models with Synthetic Motion Data for Motion Risk Prediction

Autonomous driving has seen significant progress, driven by extensive real-world data. However, in long-tail scenarios, accurately predicting the safety of the ego vehicle's future motion remains a major challenge due to uncertainties in dynamic environments and limitations in data coverage. In this work, we aim to explore whether it is possible to enhance the motion risk prediction capabilities of Vision-Language Models (VLM) by synthesizing high-risk motion data. Specifically, we introduce a Bird's-Eye View (BEV) based motion simulation method to model risks from three aspects: the ego-vehicle, other vehicles, and the environment. This allows us to synthesize plug-and-play, high-risk motion data suitable for VLM training, which we call DriveMRP-10K. Furthermore, we design a VLM-agnostic motion risk estimation framework, named DriveMRP-Agent. This framework incorporates a novel information injection strategy for global context, ego-vehicle perspective, and trajectory projection, enabling VLMs to effectively reason about the spatial relationships between motion waypoints and the environment. Extensive experiments demonstrate that by fine-tuning with DriveMRP-10K, our DriveMRP-Agent framework can significantly improve the motion risk prediction performance of multiple VLM baselines, with the accident recognition accuracy soaring from 27.13% to 88.03%. Moreover, when tested via zero-shot evaluation on an in-house real-world high-risk motion dataset, DriveMRP-Agent achieves a significant performance leap, boosting the accuracy from base_model's 29.42% to 68.50%, which showcases the strong generalization capabilities of our method in real-world scenarios.

cs.CV↗

arXiv2025

Remember Past, Anticipate Future: Learning Continual Multimodal Misinformation Detectors

Nowadays, misinformation articles, especially multimodal ones, are widely spread on social media platforms and cause serious negative effects. To control their propagation, Multimodal Misinformation Detection (MMD) becomes an active topic in the community to automatically identify misinformation. Previous MMD methods focus on supervising detectors by collecting offline data. However, in real-world scenarios, new events always continually emerge, making MMD models trained on offline data consistently outdated and ineffective. To address this issue, training MMD models under online data streams is an alternative, inducing an emerging task named continual MMD. Unfortunately, it is hindered by two major challenges. First, training on new data consistently decreases the detection performance on past data, named past knowledge forgetting. Second, the social environment constantly evolves over time, affecting the generalization on future data. To alleviate these challenges, we propose to remember past knowledge by isolating interference between event-specific parameters with a Dirichlet process-based mixture-of-expert structure, and anticipate future environmental distributions by learning a continuous-time dynamics model. Accordingly, we induce a new continual MMD method DAEDCMD. Extensive experiments demonstrate that DAEDCMD can consistently and significantly outperform the compared methods, including six MMD baselines and three continual learning methods.

cs.CL↗

arXiv2025

Label-Free Long-Horizon 3D UAV Trajectory Prediction via Motion-Aligned RGB and Event Cues

The widespread use of consumer drones has introduced serious challenges for airspace security and public safety. Their high agility and unpredictable motion make drones difficult to track and intercept. While existing methods focus on detecting current positions, many counter-drone strategies rely on forecasting future trajectories and thus require more than reactive detection to be effective. To address this critical gap, we propose an unsupervised vision-based method for predicting the three-dimensional trajectories of drones. Our approach first uses an unsupervised technique to extract drone trajectories from raw LiDAR point clouds, then aligns these trajectories with camera images through motion consistency to generate reliable pseudo-labels. We then combine kinematic estimation with a visual Mamba neural network in a self-supervised manner to predict future drone trajectories. We evaluate our method on the challenging MMAUD dataset, including the V2 sequences that feature wide-field-of-view multimodal sensors and dynamic UAV motion in urban scenes. Extensive experiments show that our framework outperforms supervised image-only and audio-visual baselines in long-horizon trajectory prediction, reducing 5-second 3D error by around 40 percent without using any manual 3D labels. The proposed system offers a cost-effective, scalable alternative for real-time counter-drone deployment. All code will be released upon acceptance to support reproducible research in the robotics community.

cs.RO↗

arXiv2025

Nonlinear Non-Hermitian Skin Effect and Skin Solitons in Temporal Photonic Feedforward Lattices

Here we report the experimental demonstration of the nonlinear non-Hermitian skin effect (NHSE) in an effective Kerr nonlinear temporal photonic lattice, where the high-power requirements and lack of tunability intrinsic to optical materials are overcome by an artificial nonlinearity arising from optoelectronic feedforward. Thanks to Kerr self-trapping, the nonlinear NHSE is demonstrated to possess much better localization strength and robustness at the preferred boundary compared to the linear case. Away from the preferred boundary, Kerr self-trapping can even inhibit NHSE-induced transport and form stable skin solitons. Harnessing the nonlinearity-controlled NHSE, we judiciously design an optical router with a flexibly tuned output port. Our findings promise great applications in robust signal transmission, routing, and processing.

physics.optics↗

arXiv2025

Genesis: Multimodal Driving Scene Generation with Spatio-Temporal and Cross-Modal Consistency

We present Genesis, a unified framework for joint generation of multi-view driving videos and LiDAR sequences with spatio-temporal and cross-modal consistency. Genesis employs a two-stage architecture that integrates a DiT-based video diffusion model with 3D-VAE encoding, and a BEV-aware LiDAR generator with NeRF-based rendering and adaptive sampling. Both modalities are directly coupled through a shared latent space, enabling coherent evolution across visual and geometric domains. To guide the generation with structured semantics, we introduce DataCrafter, a captioning module built on vision-language models that provides scene-level and instance-level supervision. Extensive experiments on the nuScenes benchmark demonstrate that Genesis achieves state-of-the-art performance across video and LiDAR metrics (FVD 16.95, FID 4.24, Chamfer 0.611), and benefits downstream tasks including segmentation and 3D detection, validating the semantic fidelity and practical utility of the generated data.

cs.CV↗

arXiv2025

Dual-view Spatio-Temporal Feature Fusion with CNN-Transformer Hybrid Network for Chinese Isolated Sign Language Recognition

Due to the emergence of many sign language datasets, isolated sign language recognition (ISLR) has made significant progress in recent years. In addition, the development of various advanced deep neural networks is another reason for this breakthrough. However, challenges remain in applying the technique in the real world. First, existing sign language datasets do not cover the whole sign vocabulary. Second, most of the sign language datasets provide only single view RGB videos, which makes it difficult to handle hand occlusions when performing ISLR. To fill this gap, this paper presents a dual-view sign language dataset for ISLR named NationalCSL-DP, which fully covers the Chinese national sign language vocabulary. The dataset consists of 134140 sign videos recorded by ten signers with respect to two vertical views, namely, the front side and the left side. Furthermore, a CNN transformer network is also proposed as a strong baseline and an extremely simple but effective fusion strategy for prediction. Extensive experiments were conducted to prove the effectiveness of the datasets as well as the baseline. The results show that the proposed fusion strategy can significantly increase the performance of the ISLR, but it is not easy for the sequence-to-sequence model, regardless of whether the early-fusion or late-fusion strategy is applied, to learn the complementary features from the sign videos of two vertical views.

cs.CV↗

arXiv2025

A Novel Fine Spectral Structure of Solar Radio Bursts with Periodic Beaded Stripes Observed by CBSm of CMP-II

A novel fine spectral structure in solar radio bursts has been discovered using the Chashan broadband solar radio spectrometer at meter wavelengths (CBSm), an instrument of the Chinese Meridian Project-Phase II (CMP-II). The structure features periodic narrow-band stripes with a typical recurrence time $< 1 $ s (occasionally reaches 8 s), often drifting from high to low frequencies and accompanied by absorptions, with trailing stripes appearing at the end of preceding ones. Some stripes exhibit periodic beaded enhancements with a periodicity of $\sim$0.1 s. The beaded stripes are reported for the first time ever. Data from the DAocheng Radio Telescope (DART) indicate a radio emission brightness temperature exceeding $10^{9}$ K, originating above brightening loops in active region AR 13664. We proposed a novel generation mechanism of the periodic stripes on the basis of the double plasma resonance (DPR) instability, and explained the beaded substructure in terms of modulation by low-frequency magnetohydrodynamic (MHD) waves. The study highlights the CBSm's capability to detect high-resolution fine spectral structures and offers novel insights into the emission mechanism and source characteristics of solar radio bursts.

astro-ph.SR↗

arXiv2025

Extending Large Vision-Language Model for Diverse Interactive Tasks in Autonomous Driving

The Large Visual-Language Models (LVLMs) have significantly advanced image understanding. Their comprehension and reasoning capabilities enable promising applications in autonomous driving scenarios. However, existing research typically focuses on front-view perspectives and partial objects within scenes, struggling to achieve comprehensive scene understanding. Meanwhile, existing LVLMs suffer from the lack of mapping relationship between 2D and 3D and insufficient integration of 3D object localization and instruction understanding. To tackle these limitations, we first introduce NuInteract, a large-scale dataset with over 1.5M multi-view image language pairs spanning dense scene captions and diverse interactive tasks. Furthermore, we propose DriveMonkey, a simple yet effective framework that seamlessly integrates LVLMs with a spatial processor using a series of learnable queries. The spatial processor, designed as a plug-and-play component, can be initialized with pre-trained 3D detectors to improve 3D perception. Our experiments show that DriveMonkey outperforms general LVLMs, especially achieving a 9.86% notable improvement on the 3D visual grounding task. The dataset and code will be released at https://github.com/zc-zhao/DriveMonkey.

cs.CV↗

arXiv2025

Robust Misinformation Detection by Visiting Potential Commonsense Conflict

The development of Internet technology has led to an increased prevalence of misinformation, causing severe negative effects across diverse domains. To mitigate this challenge, Misinformation Detection (MD), aiming to detect online misinformation automatically, emerges as a rapidly growing research topic in the community. In this paper, we propose a novel plug-and-play augmentation method for the MD task, namely Misinformation Detection with Potential Commonsense Conflict (MD-PCC). We take inspiration from the prior studies indicating that fake articles are more likely to involve commonsense conflict. Accordingly, we construct commonsense expressions for articles, serving to express potential commonsense conflicts inferred by the difference between extracted commonsense triplet and golden ones inferred by the well-established commonsense reasoning tool COMET. These expressions are then specified for each article as augmentation. Any specific MD methods can be then trained on those commonsense-augmented articles. Besides, we also collect a novel commonsense-oriented dataset named CoMis, whose all fake articles are caused by commonsense conflict. We integrate MD-PCC with various existing MD backbones and compare them across both 4 public benchmark datasets and CoMis. Empirical results demonstrate that MD-PCC can consistently outperform the existing MD baselines.

cs.CL↗

arXiv2025

JANC: A cost-effective, differentiable compressible reacting flow solver featured with JAX-based adaptive mesh refinement

The compressible reacting flow numerical solver is an essential tool in the study of combustion, energy disciplines, as well as in the design of industrial power and propulsion devices. We have established the first JAX-based block-structured adaptive mesh refinement (AMR) framework, called JAX-AMR, and then developed a fully-differentiable solver for compressible reacting flows, named JANC. JANC is implemented in Python and features automatic differentiation capabilities, enabling an efficient integration of the solver with machine learning. Furthermore, benefited by multiple acceleration features such as XLA-powered JIT compilation, GPU/TPU computing, parallel computing, and AMR, the computational efficiency of JANC has been significantly improved. In a comparative test of a two-dimensional detonation tube case, the computational cost of the JANC core solver, running on a single A100 GPU, was reduced to 1% of that of OpenFOAM, which was parallelized across 384 CPU cores. When the AMR method is enabled for both solvers, JANC's computational cost can be reduced to 1-2% of that of OpenFOAM. The core solver of JANC has also been tested for parallel computation on a 4-card A100 setup, demonstrating its convenient and efficient parallel computing capability. JANC also shows strong compatibility with machine learning by combining adjoint optimization to make the whole dynamic trajectory efficiently differentiable. JANC provides a new generation of high-performance, cost-effective, and high-precision solver framework for large-scale numerical simulations of compressible reacting flows and related machine learning research. Now, the source codes have been available under the MIT license at https://github.com/JA4S/JAX-AMR and https://github.com/JA4S/JANC.

physics.flu-dyn↗

arXiv2025

NormalCrafter: Learning Temporally Consistent Normals from Video Diffusion Priors

Surface normal estimation serves as a cornerstone for a spectrum of computer vision applications. While numerous efforts have been devoted to static image scenarios, ensuring temporal coherence in video-based normal estimation remains a formidable challenge. Instead of merely augmenting existing methods with temporal components, we present NormalCrafter to leverage the inherent temporal priors of video diffusion models. To secure high-fidelity normal estimation across sequences, we propose Semantic Feature Regularization (SFR), which aligns diffusion features with semantic cues, encouraging the model to concentrate on the intrinsic semantics of the scene. Moreover, we introduce a two-stage training protocol that leverages both latent and pixel space learning to preserve spatial accuracy while maintaining long temporal context. Extensive evaluations demonstrate the efficacy of our method, showcasing a superior performance in generating temporally consistent normal sequences with intricate details from diverse videos.

cs.CV↗

arXiv2025

ChineseSafe: A Chinese Benchmark for Evaluating Safety in Large Language Models

With the rapid development of Large language models (LLMs), understanding the capabilities of LLMs in identifying unsafe content has become increasingly important. While previous works have introduced several benchmarks to evaluate the safety risk of LLMs, the community still has a limited understanding of current LLMs' capability to recognize illegal and unsafe content in Chinese contexts. In this work, we present a Chinese safety benchmark (ChineseSafe) to facilitate research on the content safety of large language models. To align with the regulations for Chinese Internet content moderation, our ChineseSafe contains 205,034 examples across 4 classes and 10 sub-classes of safety issues. For Chinese contexts, we add several special types of illegal content: political sensitivity, pornography, and variant/homophonic words. Moreover, we employ two methods to evaluate the legal risks of popular LLMs, including open-sourced models and APIs. The results reveal that many LLMs exhibit vulnerability to certain types of safety issues, leading to legal risks in China. Our work provides a guideline for developers and researchers to facilitate the safety of LLMs. Our results are also available at https://huggingface.co/spaces/SUSTech/ChineseSafe-Benchmark. Additionally, we release a test set comprising 200,000 examples, which is publicly accessible at https://huggingface.co/datasets/SUSTech/ChineseSafe.

cs.CL↗

arXiv2025

CoGen: 3D Consistent Video Generation via Adaptive Conditioning for Autonomous Driving

Recent progress in driving video generation has shown significant potential for enhancing self-driving systems by providing scalable and controllable training data. Although pretrained state-of-the-art generation models, guided by 2D layout conditions (e.g., HD maps and bounding boxes), can produce photorealistic driving videos, achieving controllable multi-view videos with high 3D consistency remains a major challenge. To tackle this, we introduce a novel spatial adaptive generation framework, CoGen, which leverages advances in 3D generation to improve performance in two key aspects: (i) To ensure 3D consistency, we first generate high-quality, controllable 3D conditions that capture the geometry of driving scenes. By replacing coarse 2D conditions with these fine-grained 3D representations, our approach significantly enhances the spatial consistency of the generated videos. (ii) Additionally, we introduce a consistency adapter module to strengthen the robustness of the model to multi-condition control. The results demonstrate that this method excels in preserving geometric fidelity and visual realism, offering a reliable video generation solution for autonomous driving.

cs.CV↗