论文检索

Chenxu Han, Sean Bin Yang, Jilin Hu

Map matching for sparse trajectories is a fundamental problem for many trajectory-based applications, e.g., traffic scheduling and traffic flow analysis. Existing methods for map matching are generally based on Hidden Markov Model (HMM) or encoder-decoder framework. However, these methods continue to face significant challenges when handling noisy or sparsely sampled GPS trajectories. To address these limitations, we propose DiffMM, an encoder–diffusion-based map matching framework that produces effective yet efficient matching results through a one-step diffusion process. We first introduce a road segment-aware trajectory encoder that jointly embeds the input trajectory and its surrounding candidate road segments into a shared latent space through an attention mechanism. Next, we propose a one step diffusion method to realize map matching through a shortcut model by leveraging the joint embedding of the trajectory and candidate road segments as conditioning context. We conduct extensive experiments on large-scale trajectory datasets, demonstrating that our approach consistently outperforms state-of-the-art map matching methods in terms of both accuracy and efficiency, particularly for sparse trajectories and complex road network topologies.

Xiaoxi Cui, Chao Zhao, Yurong Cheng, Xiangmin Zhou

Sequential recommendation (SR) aims to predict users' next action based on their historical behavior, and is widely adopted by a number of platforms. The performance of SR models relies on rich interaction data. However, in real-world scenarios, many users only have a few historical interactions, leading to the problem of data sparsity. Data sparsity not only leads to model overfitting on sparse sequences, but also hinders the model’s ability to capture the underlying hierarchy of user intents. This results in misinterpreting the user's true intents and recommending irrelevant items. Existing data augmentation methods attempt to mitigate overfitting by generating relevant and varied data. However, they overlook the problem of reconstructing the user's intent hierarchy, which is lost in sparse data. Consequently, the augmented data often fails to align with the user's true intents, potentially leading to misguided recommendations. To address this, we propose the Adaptive Diffusion Augmentation for Recommendation (ADARec) framework. Critically, instead of using a diffusion model as a black-box generator, we use its entire step-wise denoising trajectory to reconstruct a user's intent hierarchy from a single sparse sequence. To ensure both efficiency and effectiveness, our framework adaptively determines the required augmentation depth for each sequence and employs a specialized mixture-of-experts architecture to decouple coarse- and fine-grained intents. Experiments show ADARec outperforms state-of-the-art methods on standard benchmarks and on sparse sequences, demonstrating its ability to reconstruct hierarchical intent representations from sparse data.

Yong Chen, Li Li, Nannan Zong, Zhihui Liu, Song-Zhi Su

Learning representations on graphs is foundational for many downstream tasks, and its synergy with diffusion models has emerged as a promising direction. However, diffusion-based methods for heterogeneous graphs remain underexplored, confronting two principal challenges: (1) The presence of noise and structural heterogeneity in graphs makes it challenging to accurately capture semantic transitions among diverse relation types. (2) The isotropic Gaussian noise used in forward diffusion fails to reflect graphs' inherent semantics and structural anisotropy. To address these, we propose ARDiff, a novel framework that integrates residual diffusion with anisotropic noise for heterogeneous graph learning. Specifically, we propose a semantic residual diffusion mechanism that progressively refines node embeddings by orchestrating transitions from low-semantic (high-noise) to high-semantic (low-noise) relational contexts, thus enabling step-wise distillation of task-relevant information. In addition, to address the limitations of conventional diffusion, we introduce an anisotropic diffusion strategy: in the forward process, noise injection is oriented by structural and semantic priors; in the denoising step, a conditional diffusion mechanism is guided by a random walk encoding, enhancing both topological consistency and semantic alignment. Extensive evaluation on heterogeneous graph datasets demonstrates that ARDiff significantly surpasses current leading methods in link prediction and node classification, setting a new paradigm and benchmark in heterogeneous graph representation learning.

Mingyang Chen, Jiawei Du, Bo Huang, Yi Wang, Xiaobo Zhang, Wei Wang

Existing core-set selection methods predominantly rely on heuristic scoring signals such as training dynamics or model uncertainty, lacking explicit modeling of data likelihood. This omission may hinder the constructed subset from capturing subtle yet critical distributional structures that underpin effective model training. In this work, we propose a novel, theoretically grounded approach that leverages diffusion models to estimate data likelihood via reconstruction deviation induced by partial reverse denoising. Specifically, we establish a formal connection between reconstruction error and data likelihood, grounded in the Evidence Lower Bound (ELBO) of Markovian diffusion processes, thereby enabling a principled, distribution-aware scoring criterion for data selection. Complementarily, we introduce an efficient information-theoretic method to identify the optimal reconstruction timestep, ensuring that the deviation provides a reliable signal indicative of underlying data likelihood. Extensive experiments on ImageNet demonstrate that reconstruction deviation offers an effective scoring criterion, consistently outperforming existing baselines across selection ratios, and closely matching full-data training using only 50% of the data. Further analysis shows that the likelihood-informed nature of our score reveals informative insights in data selection, shedding light on the interplay between data distributional characteristics and model learning preferences.

Ruobing Wang, Xin Li, Yangchuan Wang, Zijian Zhang, Mingzhong Wang

Branch-and-bound (B&B) is a fundamental algorithmic framework for solving Mixed-Integer Linear Programming (MILP) problems, where branching decisions critically affect solver efficiency. Recent learning-based methods apply imitation learning to select branching variables, but their deterministic predictions limit exploration and generalization. In this paper, we propose a novel framework that formulates branching variable selection as a conditional generative process, exploring deep-level decision features. Our approach leverages diffusion models to enable diverse and exploratory branching score generation, while consistency modeling distills this process into efficient one-step inference conditioned on the B&B state. This mode allows our method to achieve both high-quality and fast branching decisions, significantly improving the overall performance of branch-and-bound solvers. Extensive experiments on challenging cross-scale and cross-category benchmarks demonstrate that our framework consistently outperforms state-of-the-art imitation learning baselines, delivering substantial improvements in solution quality, computational efficiency, and inference speed.

Ruben Becker, Bojana Kodric, Cosimo Vinci

We study a general framework of optimization with the aim to compute fair solutions in settings with a set of agents whose valuations are combined using an aggregation function. The strength of our framework lies (1) in its generality and (2) in the fact that we leverage the power of ex-ante fairness, a concept that has recently gained much attention in the scope of fair allocation and fairness in AI in general. More precisely, in our setting there are n set functions f₁, …, fₙ (e.g., the valuation functions of n agents) that are combined using an aggregation function g (e.g., the minimum, Nash social welfare, p-norm). The power of ex-ante fairness is obtained by allowing as a feasible solution not simply a finite set S, but instead a distribution Π over feasible sets. The goal in our setting is then to find a probability distribution p in Π that maximizes the value resulting from aggregating (using g) the n expected values of the functions f₁, …, fₙ obtained when sampling a set S according to the distribution p. We stress that this is different from maximizing the expected value of g (ex-post fairness) and typically allows for much fairer solutions. We give three different greedy algorithms for three different settings of this framework and prove that they achieve constant approximation guarantees under certain realistic assumptions. For some of the settings, we show that these approximation guarantees are tight. Specific scenarios that can be modelled using our framework include fair information diffusion in social networks, fair submodular matching problems, and ex-ante versions of item assignment problems.

Yang Zou, Xingyue Zhu, Kaiqi Han, Jun Ma, Xingyuan Li, Zhiying Jiang, Jinyuan Liu

Infrared video has been of great interest in visual tasks under challenging environments, but often suffers from severe atmospheric turbulence and compression degradation. Existing video super-resolution (VSR) methods either neglect the inherent modality gap between infrared and visible images or fail to restore turbulence-induced distortions. Directly cascading turbulence mitigation (TM) algorithms with VSR methods leads to error propagation and accumulation due to the decoupled modeling of degradation between turbulence and resolution. We introduce HATIR, a Heat-Aware Diffusion for Turbulent InfraRed Video Super-Resolution, which injects heat-aware deformation priors into the diffusion sampling path to jointly model the inverse process of turbulent degradation and structural detail loss. Specifically, HATIR constructs a Phasor-Guided Flow Estimator, rooted in the physical principle that thermally active regions exhibit consistent phasor responses over time, enabling reliable turbulence-aware flow to guide the reverse diffusion process. To ensure the fidelity of structural recovery under nonuniform distortions, a Turbulence-Aware Decoder is proposed to selectively suppress unstable temporal cues and enhance edge-aware feature aggregation via turbulence gating and structure-aware attention. We built FLIR-IVSR, the first dataset for turbulent infrared VSR, comprising paired LR-HR sequences from a FLIR T1050sc camera (1024 X 768) spanning 640 diverse scenes with varying camera and object motion conditions. This encourages future research in infrared VSR.

Ziyang Zong, Guanying Chen, Zhaohuan Zhan, Fengcheng Yu, Guang Tan

This paper proposes a two-stage text-to-floorplan generation framework that combines the reasoning capability of Large Language Models (LLMs) with the generative power of diffusion models. In the first stage, we leverage a Chain-of-Thought (CoT) prompting strategy to guide an LLM in generating an initial layout, Layout-Init, from natural language descriptions, which ensures a user-friendly and intuitive design process. However, Layout-Init may lack precise geometric alignment and fine-grained structural details due to the inherent limitations of LLMs. To address this, in the second stage we propose a Dual-Noise Prior-Preserved Diffusion (DNPP-Diffusion) model to refine Layout-Init into a final floorplan that better adheres to physical constraints and user requirements. By combining LLMs and a dedicated refining model, our approach is able to generate high-quality floorplans without requiring large-scale domain-specific training data. Experimental results demonstrate its advantages in comparison with state of the art methods, and validate its effectiveness in home design applications.

Jian Zhu, Zhengyu Jia, Tian Gao, Jiaxin Deng, Shidi Li, Lang Zhang, Fu Liu, Peng Jia, Xianpeng Lang

Advanced end-to-end autonomous driving systems predict other vehicles' motions and plan ego vehicle's trajectory. The world model that can foresee the outcome of the trajectory has been used to evaluate the end-to-end autonomous driving system. However, existing world models predominantly emphasize the trajectory of the ego vehicle and leave other vehicles uncontrollable. This limitation hinders their ability to realistically simulate the interaction between the ego vehicle and the driving scenario. In addition, it remains a challenge to match multiple trajectories with each vehicle in the video to control the video generation. To address above issues, a driving World Model named EOT-WM is proposed in this paper, unifying Ego-Other vehicle Trajectories in videos. Specifically, we first project ego and other vehicle trajectories in the BEV space into the image coordinate to match each trajectory with its corresponding vehicle in the video. Then, trajectory videos are encoded by the Spatial-Temporal Variational Auto Encoder to align with driving video latents spatially and temporally in the unified visual space. A trajectory-injected diffusion Transformer is further designed to denoise the noisy video latents for video generation with the guidance of ego-other vehicle trajectories. In addition, we propose a metric based on control latent similarity to evaluate the controllability of trajectories. Extensive experiments are conducted on the nuScenes dataset, and the proposed model outperforms the state-of-the-art method by 30% in FID and 55% in FVD. The model can also predict unseen driving scenes with self-produced trajectories.

Aihua Zhu, Rui Su, Qinglin Zhao, Li Feng, Meng Shen, Shibo He

Diffusion probabilistic models have set a new standard for generative fidelity but are hindered by a slow iterative sampling process. A powerful training-free strategy to accelerate this process is Schedule Optimization, which aims to find an optimal distribution of timesteps for a fixed and small Number of Function Evaluations (NFE) to maximize sample quality. To this end, a successful schedule optimization method must adhere to four core principles: effectiveness, adaptivity, practical robustness, and computational efficiency. However, existing paradigms struggle to satisfy these principles simultaneously, motivating the need for a more advanced solution. To overcome these limitations, we propose the Hierarchical-Schedule-Optimizer (HSO), a novel and efficient bi-level optimization framework. HSO reframes the search for a globally optimal schedule into a more tractable problem by iteratively alternating between two synergistic levels: an upper-level global search for an optimal initialization strategy and a lower-level local optimization for schedule refinement. This process is guided by two key innovations: the Midpoint Error Proxy (MEP), a solver-agnostic and numerically stable objective for effective local optimization, and the Spacing-Penalized Fitness (SPF) function, which ensures practical robustness by penalizing pathologically close timesteps. Extensive experiments show that HSO sets a new state-of-the-art for training-free sampling in the extremely low-NFE regime. For instance, with an NFE of just 5, HSO achieves a remarkable FID of 11.94 on LAION-Aesthetics with Stable Diffusion v2.1. Crucially, this level of performance is attained not through costly retraining, but with a one-time optimization cost of less than 8 seconds, presenting a highly practical and efficient paradigm for diffusion model acceleration.

Zhenyu Zhou, Defang Chen, Can Wang, Chun Chen, Siwei Lyu

Text-to-image diffusion models are capable of generating high-quality images, but suboptimal pre-trained text representations often result in these images failing to align closely with the given text prompts. Classifier-free guidance (CFG) is a popular and effective technique for improving text-image alignment in the generative process. However, CFG introduces significant computational overhead. In this paper, we present DIstilling CFG by sharpening text Embeddings (DICE) that replaces CFG in the sampling process with half the computational complexity while maintaining similar generation quality. DICE distills a CFG-based text-to-image diffusion model into a CFG-free version by refining text embeddings to replicate CFG-based directions. In this way, we avoid the computational drawbacks of CFG, enabling high-quality, well-aligned image generation at a fast sampling speed. Furthermore, examining the enhancement pattern, we identify the underlying mechanism of DICE that sharpens specific components of text embeddings to preserve semantic information while enhancing fine-grained details. Extensive experiments on multiple Stable Diffusion v1.5 variants, SDXL, and PixArt-\alpha demonstrate the effectiveness of our method.

Zanwei Zhou, Taoran Yi, Jiemin Fang, Chen Yang, Lingxi Xie, Xinggang Wang, Wei Shen, Qi Tian

Flow-based 3D generation models typically require dozens of sampling steps during inference. Though few-step distillation methods, particularly Consistency Models (CMs), have achieved substantial advancements in accelerating 2D diffusion models, they remain under-explored for more complex 3D generation tasks. In this study, we propose a novel framework, MDT-dist, for few-step 3D flow distillation. Our approach is built upon a primary objective: distilling the pretrained model to learn the Marginal-Data Transport. Directly learning this objective needs to integrate the velocity fields, while this integral is intractable to be implemented. Therefore, we propose two optimizable objectives, Velocity Matching (VM) and Velocity Distillation (VD), to equivalently convert the optimization target from the transport level to the velocity and the distribution level respectively. Velocity Matching (VM) learns to stably match the velocity fields between the student and the teacher, but inevitably provides biased gradient estimates. Velocity Distillation (VD) further enhances the optimization process by leveraging the learned velocity fields to perform probability density distillation. When evaluated on the pioneer 3D generation framework TRELLIS, our method reduces sampling steps of each flow transformer from 25 to 1–2, achieving 0.68s (1 step x2) and 0.94s (2 steps x2) latency with 9.0x and 6.5x speedup on A800, while preserving high visual and geometric fidelity. Experiments demonstrate that our method significantly outperforms existing CM distillation methods, and enables TRELLIS to achieve superior performance in few-step 3D generation.

Yuhang Zhou, Yanxiang Zhao, Zhongyun Hua, Zhipu Liu, Zhaoquan Gu, Qing Liao, Leo Yu Zhang

Person re-identification (ReID) is a fundamental task in many real-world applications such as pedestrian trajectory tracking. However, advanced deep learning-based ReID models are highly susceptible to adversarial attacks, where imperceptible perturbations to pedestrian images can cause entirely incorrect predictions, posing significant security threats. Although numerous adversarial defense strategies have been proposed for classification tasks, their extension to metric learning tasks such as person ReID remains relatively unexplored. Moreover, the several existing defenses for person ReID fail to address the inherent unique challenges of adversarially robust ReID. In this paper, we systematically identify the challenges of adversarial defense in person ReID into two key issues: model bias and composite generalization requirements. To address them, we propose a debiased dual-invariant defense framework composed of two main phases. In the data balancing phase, we mitigate model bias using a diffusion-model-based data resampling strategy that promotes fairness and diversity in training data. In the bi-adversarial self-meta defense phase, we introduce a novel metric adversarial training approach incorporating farthest negative extension softening to overcome the robustness degradation caused by the absence of classifier. Additionally, we introduce an adversarially-enhanced self-meta mechanism to achieve dual-generalization for both unseen identities and unseen attack types. Experiments demonstrate that our method significantly outperforms existing state-of-the-art defenses.

Yucheng Zhou, Jihai Zhang, Guanjie Chen, Jianbing Shen, Yu Cheng

Video generation using Large Language Models (LLMs) has shown promising potential, effectively leveraging the extensive LLM infrastructure to provide a unified framework for multimodal understanding and content generation. However, these methods face critical challenges, i.e., token redundancy and inefficiencies arising from long sequences, which constrain their performance and efficiency compared to diffusion-based approaches. In this study, we investigate the impact of token redundancy in LLM-based video generation by information-theoretic analysis and propose Vision Representation Compression (VRC), a novel framework designed to achieve more in both performance and efficiency with less video token representations. VRC introduces learnable representation compressor and decompressor to compress video token representations, enabling autoregressive next-sequence prediction in a compact latent space. Our approach reduces redundancy, shortens token sequences, and improves model's ability to capture underlying video structures. Our experiments demonstrate that VRC reduces token sequence lengths by a factor of 4, achieving more than 9~14x acceleration in inference while maintaining performance comparable to state-of-the-art video generation models. VRC not only accelerates the inference but also significantly reduces memory requirements during both model training and inference.

Mingxuan Zhou, Yirui Shen, Shuang Li, Jing Geng, Yutang Zhang, Shuigen Wang

The practical deployment of infrared imaging is hindered by its inherent output of low-resolution (LR) images. While the super-resolution (SR) technique is a promising remedy, we discover two major challenges concerning infrared image SR: preserving accurate thermal distributions, which are fundamental to infrared imaging, and addressing the ambiguity of high-frequency elements compared to visible images. To tackle these issues, we propose ThesIS, a tailored framework that utilizes Thermal-Physics guidance and dynamic high-frequency amplification for Infrared image Super-resolution to produce high-resolution (HR) images with accurate physical properties and delicate visual details. Specifically, Thermal Regularization is introduced to reconstruct the accurate thermal radiation distribution via the introduced Infrared Radiation Intensity Alignment Loss, mitigating the adverse effects of complex degradations while conducting initial upscaling. Additionally, we design a guidance mechanism to counter the randomness of the diffusion model, further refining the preservation of physical information. The proposed Dynamic High-Frequency Amplification effectively strengthens the ambiguous high-frequency information present in infrared images, leading to improved texture details and superior visual quality. Extensive experiments demonstrate that ThesIS successfully recovers accurate thermal information while delivering visually satisfying results with state-of-the-art performance. Furthermore, we introduce the InfraredSR dataset, which comprises 39,833 images at a resolution of 512 × 512, hoping to advance research in this field.

Jun Zhou, Chi Xu, Kaifeng Tang, Yuting Ge, Tingrui Guo, Li Cheng

Estimating the 3D poses of hands and objects from a single RGB image is a fundamental yet challenging problem, with broad applications in augmented reality and human-computer interaction. Existing methods largely rely on visual cues alone, often producing results that violate physical constraints such as interpenetration or non-contact. Recent efforts to incorporate physics reasoning typically depend on post-optimization or non-differentiable physics engines, which compromise visual consistency and end-to-end trainability. To overcome these limitations, we propose a novel framework that jointly integrates visual and physical cues for hand-object pose estimation. This integration is achieved through two key ideas: 1) joint visual-physical cue learning: The model is trained to extract 2D visual cues and 3D physical cues, thereby enabling more comprehensive representation learning for hand-object interactions; 2) candidate pose aggregation: A novel refinement process that aggregates multiple diffusion-generated candidate poses by leveraging both visual and physical predictions, yielding a final estimate that is visually consistent and physically plausible. Extensive experiments demonstrate that our method significantly outperforms existing state-of-the-art approaches in both pose accuracy and physical plausibility.

Jin Zhou, Yi Zhou, Hongliang Yang, Pengfei Xu, Hui Huang

In the field of sketch generation, raster-format trained models often produce non-stroke artifacts, while vector-format trained models typically lack a holistic understanding of sketches, leading to compromised recognizability. Moreover, existing methods struggle to extract common features from similar elements (e.g., eyes of animals) appearing at varying positions across sketches. To address these challenges, we propose StrokeFusion, a two-stage framework for vector sketch generation. It contains a dual-modal sketch feature learning network that maps strokes into a high-quality latent space. This network decomposes sketches into normalized strokes and jointly encodes stroke sequences with Unsigned Distance Function (UDF) maps, representing sketches as sets of stroke feature vectors. Building upon this representation, our framework exploits a stroke-level latent diffusion model that simultaneously adjusts stroke position, scale, and trajectory during generation. This enables high-fidelity stroke generation while supporting stroke interpolation editing. Extensive experiments across multiple sketch datasets, demonstrate that our framework outperforms state-of-the-art techniques, validating its effectiveness in preserving structural integrity and semantic features.

Haoyi Zhong, Fang-Lue Zhang, Andrew Chalmers, Taehyun Rhee

While instruction-based image editing is emerging, extending it to 360° panorama introduces additional challenges. Existing methods often produce implausible results in both equirectangular projections (ERP) and perspective views. To address these limitations, we propose SE360, a novel framework for multi-condition guided object editing in 360° panoramas. At its core is a novel coarse-to-fine autonomous data generation pipeline without manual intervention. This pipeline leverages a Vision-Language Model (VLM) and adaptive projection adjustment for hierarchical analysis, ensuring the holistic segmentation of objects and their physical context. The resulting data pairs are both semantically meaningful and geometrically consistent, even when sourced from unlabeled panoramas. Furthermore, we introduce a cost-effective, two-stage data refinement strategy to improve data realism and mitigate model overfitting to erasing artifacts. Based on the constructed dataset, we train a Transformer-based diffusion model to allow flexible object editing guided by text, mask, or reference image in 360° panoramas. Our experiments demonstrate that our method outperforms existing methods in both visual quality and semantic accuracy.

Wenli Zheng, Huiyuan Fu, Zekai Xu, Xin Wang, Huadong Ma

High-scale image super-resolution (SR) has become increasingly important with the rapid growth of mobile devices and high-resolution displays. However, current SR methods primarily focus on lower scales and generalize poorly to high-scale scenarios due to severe information loss and complex real-world degradations. In this paper, we propose a novel Selective Diffusion Distillation (SDD) framework for real-world high-scale SR, which distills reliable knowledge from a low-scale diffusion teacher to a high-scale student. Specifically, considering severe information loss in high-scale inputs, directly distilling from low-scale models may result in feature misalignment. To address this, we introduce a Degradation-aware Metric Learning (DML) approach to align feature distributions across different degradation levels. In addition, since the diffusion-based teacher may hallucinate artifacts in ambiguous regions, blindly imitating these unreliable outputs can degrade the student’s fidelity. To tackle this, we propose a Region-aware Selective Distillation (RSD) strategy to filter out uncertain predictions and adaptively supervise only on reliable areas. To evaluate the effectiveness of our method, we introduce Real-UltraSR, a new real-world benchmark that contains diverse high-scale LR-HR pairs, including x8, x10, x12, and x14. Extensive experiments demonstrate that our SDD framework achieves state-of-the-art performance across multiple benchmarks.

Shikang Zheng, Liang Feng, Xinyu Wang, Qinming Zhou, Peiliang Cai, Chang Zou, Jiacheng Liu, Yuqi Lin, Junjie Chen, Yue Ma 等

Diffusion Transformers (DiTs) have demonstrated exceptional performance in high-fidelity image and video generation. To reduce their substantial computational costs, feature caching techniques have been proposed to accelerate inference by reusing hidden representations from previous timesteps. However, current methods often struggle to maintain generation quality at high acceleration ratios, where prediction errors increase sharply due to the inherent instability of long-step forecasting. In this work, we adopt an ordinary differential equation (ODE) perspective on the hidden-feature sequence, modeling layer representations along the trajectory as a feature-ODE. We attribute the degradation of existing caching strategies to their inability to robustly integrate historical features under large skipping intervals. To address this, we propose FoCa (Forecast-then-Calibrate), which treats feature caching as a feature-ODE solving problem. Extensive experiments on image, video generation, and super-resolution tasks demonstrate the effectiveness of FoCa, especially under aggressive acceleration. Without additional training, FoCa achieves near-lossless speedups of 5.50× on FLUX, 6.45× on HunyuanVideo, 3.17× on Inf-DiT, and maintains high quality with a 4.53× speedup on DiT.