arXiv subjects

Guangyuan Li

Publications and source records attributed to Guangyuan Li.

At least 19 recordsLinked to original sources

arXiv2026

FutureWorld: A Plug-and-Play Causal Next-State Predictor for Interactive Video World Exploration

Interactive video world models aim to follow user actions while maintaining scene consistency over extended interactions. History-based methods reuse generated latents to preserve scene content, but three limitations of this reuse contribute to error recirculation. 1) The latent generated at each interaction is retained as history, potentially containing blurred details, geometric distortions, or semantic drift. 2) Although retrieved historical content matches the current scene, it can still contain errors that are carried into the next chunk. 3) New outputs then become history for later interactions, potentially reinforcing the same errors through repeated reuse. To limit this cycle, we aim to predict the future scene state from the current generated latent and the next action before generating the next interaction's video chunk. This future-state prior provides guidance for the next chunk, with the aim of reducing the errors carried forward from history. To this end, we propose a plug-and-play Causal Next-Chunk Predictor (C-NCP) with cross-model adaptability and prediction-guided modulation. We integrate C-NCP into four base generators and evaluate the augmented models across three benchmarks for interactive long-video generation and action-conditioned world modeling. Extensive experiments demonstrate reduced error accumulation and improved long-horizon consistency across diverse generation settings. The project page is at https://GuangYuanKK.github.io/futureworld.

cs.CV↗

arXiv2026

GradSplat: Gradients Inside the Forward Pass for Feed-Forward 3D Gaussian Splatting

Feed-forward 3D Gaussian Splatting predicts a scene from a few images, but our predictor is less accurate than its own Gaussians optimized per scene on those images. This gap lies mostly in the shape of the Gaussians, their scale and rotation, since optimizing the shape alone recovers almost all of it and opacity and color recover less. Gradient steps on the input views correct this shape, and the model holds these views at inference. Log-anisotropy correction changes the axis ratios of each Gaussian in log space at constant volume, and in-forward gradient refinement renders the Gaussians into the input views and applies the gradient of the photometric loss with a learned recurrent update, first to opacity and color and then to the shape. GradSplat is a feed-forward model that refines its own Gaussians by learned gradient steps inside its forward pass, with no target view. On both complete six-view RealEstate10K protocols, it has the best PSNR and SSIM among nine methods and keeps its accuracy from $6$ to $42$ input views. Without fine-tuning, it is the most accurate on DL3DV and has the best PSNR and SSIM on ScanNet++. The code will be released at https://github.com/Jingong-Chen/GradSplat/.

cs.CV↗

arXiv2026

Forget the Data, Keep the Capability: Switchable Unlearning for Retrieval

Privacy compliance requires re-identification models to selectively forget individuals while preserving their ability to recognize others. For this, we present MoME, a switchable framework that can adapt pre-trained backbones for privacy compliance, allowing the same model to switch between Full mode for standard inference and Core mode for unlearned inference. MoME splits one backbone into core parameters that preserve retrieval and zero-initialized parasitic parameters that carry the full-data knowledge, and training them under gradient isolation so the core never receives a forget-set gradient. Disabling the branch yields Core, which satisfies a request to delete a person's training data, and enabling it yields Full, which retains what the deleted data taught, so honoring the request is a flag flip rather than a retraining run. We prove Core independent of the forget set up to an audited set of coupling channels and introduce the rank-based Forgetting-Rank statistic to audit that independence in a trained model. On Market-1501, VeRi-776 and RegDB, Core reads at a retrained reference under three membership attacks while retrieving above it, and the retrieval advantage carries to ResNet and CLIP backbones. Code and weights are available at https://github.com/Jingong-Chen/MoME.

cs.CV↗

arXiv2026

Readout Blindness: VLM Scores Miss the Spatial Direction Their Frozen Encoders Retain

CLIP-like vision-language models remain a cornerstone of multimodal systems, yet their scores stay near chance on directed spatial relations, such as whether one object is left of another. We call this failure readout blindness and analyze, theoretically and empirically, why deployed scores miss the direction: when scoring rules treat the subject and object symmetrically, direction cancels regardless of encoder training. Guided by this analysis, we introduce Antisymmetric Displacement Readout (ADR), which aligns caption words with image patches in the frozen features and scores each relation by the signed displacement between matched object centroids. Notably, ADR succeeds without additional training or learned parameters, thereby demonstrating that directional information remains in the frozen encoder. However, text and world priors can inflate accuracy, so we further introduce prior deflation, which measures the benefit of the image-text pairing as the grounded gain over a null that pairs each item with an unrelated image. Extensive experiments across encoder families show that ADR substantially improves over deployed scores, which remain near chance on most direction-balanced sets even for fine-tuned encoders. Compared with more complex readouts, ADR outperforms the evaluated MLLM likelihood readouts and is competitive with their chat inference at a small fraction of the computation. These results support our claim that directional information can be recovered from frozen features by an appropriate readout. Our implementation and evaluation kit will be publicly available.

cs.CV↗

arXiv2026

Are LLMs Ready to Assist Physicians? PhysAssistBench for Interactive Doctor-Patient-EHR Assistance

The most plausible near-term role of medical LLMs is to assist rather than replace physicians, yet current evaluations often test isolated capabilities: clinical knowledge, EHR system interaction, or patient communication. Physician assistance instead requires coordinating these capabilities within the same interaction, where physicians issue underspecified requests, patients describe symptoms ambiguously, and EHR systems demand precise tool use. We introduce PhysAssistBench, a benchmark for interactive doctor-patient-EHR assistance. Built from real MIMIC-IV cases, PhysAssistBench uses a scalable pipeline to construct agentic patients: interactive, record-grounded agents that turn static EHR records into multi-turn clinical scenarios while preserving clinical factuality. PhysAssistBench provides a curated bilingual evaluation set of 1,296 manually reviewed and physician-validated turns. Experiments with leading LLMs show that current models remain unreliable in this setting, which exposes a key bottleneck for clinical LLMs: reliable assistance requires coordination across knowledge, communication, and systems, not isolated gains in any of them.

cs.CL↗

arXiv2026

Object-Centric Vision Token Pruning for Vision Language Models

In Vision Language Models (VLMs), vision tokens are quantity-heavy yet information-dispersed compared with language tokens, thus consume too much unnecessary computation. Pruning redundant vision tokens for high VLM inference efficiency has been continuously studied but all existing methods resort to indirect and non-guaranteed ways. We propose OC-VTP, a direct and guaranteed approach to select the most representative vision tokens for high-efficiency yet accuracy-preserving VLM inference. Our OC-VTP requires merely light-weight pre-training of a small object-centric vision token pruner, which can then be inserted into existing VLMs, without fine-tuning of any models on any datasets. It is gauranteed that the most representative vision tokens are kept by minimizing the error in reconstructing the original unpruned tokens from the selected ones. Across any vision pruning ratios, i.e., inference efficiency, our OC-VTP consistently helps mainstream VLMs to preserve the highest inference accuracy. Our pruning also demonstrates interesting interpretability. Our codes are available at https://github.com/GarryLarry010131/OC-VTP.

cs.CV↗

arXiv2026

SmartPhotoCrafter: Unified Reasoning, Generation and Optimization for Automatic Photographic Image Editing

Traditional photographic image editing typically requires users to possess sufficient aesthetic understanding to provide appropriate instructions for adjusting image quality and camera parameters. However, this paradigm relies on explicit human instruction of aesthetic intent, which is often ambiguous, incomplete, or inaccessible to non-expert users. In this work, we propose SmartPhotoCrafter, an automatic photographic image editing method which formulates image editing as a tightly coupled reasoning-to-generation process. The proposed model first performs image quality comprehension and identifies deficiencies by the Image Critic module, and then the Photographic Artist module realizes targeted edits to enhance image appeal, eliminating the need for explicit human instructions. A multi-stage training pipeline is adopted: (i) Foundation pretraining to establish basic aesthetic understanding and editing capabilities, (ii) Adaptation with reasoning-guided multi-edit supervision to incorporate rich semantic guidance, and (iii) Coordinated reasoning-to generation reinforcement learning to jointly optimize reasoning and generation. During training, SmartPhotoCrafter emphasizes photo-realistic image generation, while supporting both image restoration and retouching tasks with consistent adherence to color- and tone-related semantics. We also construct a stage-specific dataset, which progressively builds reasoning and controllable generation, effective cross-module collaboration, and ultimately high-quality photographic enhancement. Experiments demonstrate that SmartPhotoCrafter outperforms existing generative models on the task of automatic photographic enhancement, achieving photo-realistic results while exhibiting higher tonal sensitivity to retouching instructions. Project page: https://github.com/vivoCameraResearch/SmartPhotoCrafter.

cs.CV↗

arXiv2026

PlaneCycle: Training-Free 2D-to-3D Lifting of Foundation Models Without Adapters

Large-scale 2D foundation models exhibit strong transferable representations, yet extending them to 3D volumetric data typically requires retraining, adapters, or architectural redesign. We introduce PlaneCycle, a training-free, adapter-free operator for architecture-agnostic 2D-to-3D lifting of foundation models. PlaneCycle reuses the original pretrained 2D backbone by cyclically distributing spatial aggregation across orthogonal HW, DW, and DH planes throughout network depth, enabling progressive 3D fusion while preserving pretrained inductive biases. The method introduces no additional parameters and is applicable to arbitrary 2D networks. Using pretrained DINOv3 models, we evaluate PlaneCycle on six 3D classification and three 3D segmentation benchmarks. Without any training, the lifted models exhibit intrinsic 3D fusion capability and, under linear probing, outperform slice-wise 2D baselines and strong 3D counterparts, approaching the performance of fully trained models. With full fine-tuning, PlaneCycle matches standard 3D architectures, highlighting its potential as a seamless and practical 2D-to-3D lifting operator. These results demonstrate that 3D capability can be unlocked from pretrained 2D foundation models without structural modification or retraining. Code is available at https://github.com/HINTLab/PlaneCycle.

cs.CV↗

arXiv2026

MagicWorld: Towards Long-Horizon Stability for Interactive Video World Exploration

Recent interactive video world model methods generate scene evolution conditioned on user instructions. Although they achieve impressive results, two key limitations remain. First, they exhibit motion drift in complex environments with multiple interacting subjects, where dynamic subjects fail to follow realistic motion patterns during scene evolution. Second, they suffer from error accumulation in long-horizon interactions, where autoregressive generation gradually drifts from earlier scene states and causes structural and semantic inconsistencies. In this paper, we propose MagicWorld, an interactive video world model built upon an autoregressive framework. To address motion drift, we incorporate a flow-guided motion preservation constraint that mitigates motion degradation in dynamic subjects, encouraging realistic motion patterns and stable interactions during scene evolution. To mitigate error accumulation in long-horizon interactions, we design two complementary strategies, including a history cache retrieval strategy and an enhanced interactive training strategy. The former reinforces historical scene states by retrieving past generations during interaction, while the latter adopts multi-shot aggregated distillation with dual-reward weighting for interactive training, enhancing long-term stability and reducing error accumulation. In addition, we construct RealWM120K, a real-world dataset with diverse city-walk videos and multimodal annotations to support dynamic perception and long-horizon world modeling. Experimental results demonstrate that MagicWorld improves motion realism and alleviates error accumulation during long-horizon interactions.

cs.CV↗

arXiv2025

MagicTryOn: Harnessing Diffusion Transformer for Garment-Preserving Video Virtual Try-on

Video Virtual Try-On (VVT) aims to synthesize garments that appear natural across consecutive video frames, capturing both their dynamics and interactions with human motion. Despite recent progress, existing VVT methods still suffer from inadequate garment fidelity and limited spatiotemporal consistency. The reasons are: (1) under-exploitation of garment information, with limited garment cues being injected, resulting in weaker fine-detail fidelity; and (2) a lack of spatiotemporal modeling, which hampers cross-frame identity consistency and causes temporal jitter and appearance drift. In this paper, we present MagicTryOn, a diffusion-transformer based framework for garment-preserving video virtual try-on. To preserve fine-grained garment details, we propose a fine-grained garment-preservation strategy that disentangles garment cues and injects these decomposed priors into the denoising process. To improve temporal garment consistency and suppress jitter, we introduce a garment-aware spatiotemporal rotary positional embedding (RoPE) that extends RoPE within full self-attention, using spatiotemporal relative positions to modulate garment tokens. We further impose a mask-aware loss during training to enhance fidelity within garment regions. Moreover, we adopt distribution-matching distillation to compress the sampling trajectory to four steps, enabling real-time inference without degrading garment fidelity. Extensive quantitative and qualitative experiments demonstrate that MagicTryOn outperforms existing methods, delivering superior garment-detail fidelity and temporal stability in unconstrained settings.

cs.CV↗

arXiv2025

Incomplete Multi-view Clustering via Diffusion Contrastive Generation

Incomplete multi-view clustering (IMVC) has garnered increasing attention in recent years due to the common issue of missing data in multi-view datasets. The primary approach to address this challenge involves recovering the missing views before applying conventional multi-view clustering methods. Although imputation-based IMVC methods have achieved significant improvements, they still encounter notable limitations: 1) heavy reliance on paired data for training the data recovery module, which is impractical in real scenarios with high missing data rates; 2) the generated data often lacks diversity and discriminability, resulting in suboptimal clustering results. To address these shortcomings, we propose a novel IMVC method called Diffusion Contrastive Generation (DCG). Motivated by the consistency between the diffusion and clustering processes, DCG learns the distribution characteristics to enhance clustering by applying forward diffusion and reverse denoising processes to intra-view data. By performing contrastive learning on a limited set of paired multi-view samples, DCG can align the generated views with the real views, facilitating accurate recovery of views across arbitrary missing view scenarios. Additionally, DCG integrates instance-level and category-level interactive learning to exploit the consistent and complementary information available in multi-view data, achieving robust and end-to-end clustering. Extensive experiments demonstrate that our method outperforms state-of-the-art approaches. The code is available at https://github.com/zhangyuanyang21/2025-AAAI-DCG.

cs.CV↗

arXiv2025

Joint Optimization of Triangle Mesh, Material, and Light from Neural Fields with Neural Radiance Cache

Traditional inverse rendering techniques are based on textured meshes, which naturally adapts to modern graphics pipelines, but costly differentiable multi-bounce Monte Carlo (MC) ray tracing poses challenges for modeling global illumination. Recently, neural fields has demonstrated impressive reconstruction quality but falls short in modeling indirect illumination. In this paper, we introduce a simple yet efficient inverse rendering framework that combines the strengths of both methods. Specifically, given pre-trained neural field representing the scene, we can obtain an initial estimate of the signed distance field (SDF) and create a Neural Radiance Cache (NRC), an enhancement over the traditional radiance cache used in real-time rendering. By using the former to initialize differentiable marching tetrahedrons (DMTet) and the latter to model indirect illumination, we can compute the global illumination via single-bounce differentiable MC ray tracing and jointly optimize the geometry, material, and light through back propagation. Experiments demonstrate that, compared to previous methods, our approach effectively prevents indirect illumination effects from being baked into materials, thus obtaining the high-quality reconstruction of triangle mesh, Physically-Based (PBR) materials, and High Dynamic Range (HDR) light probe.

cs.GR↗

arXiv2025

DyArtbank: Diverse Artistic Style Transfer via Pre-trained Stable Diffusion and Dynamic Style Prompt Artbank

Artistic style transfer aims to transfer the learned style onto an arbitrary content image. However, most existing style transfer methods can only render consistent artistic stylized images, making it difficult for users to get enough stylized images to enjoy. To solve this issue, we propose a novel artistic style transfer framework called DyArtbank, which can generate diverse and highly realistic artistic stylized images. Specifically, we introduce a Dynamic Style Prompt ArtBank (DSPA), a set of learnable parameters. It can learn and store the style information from the collection of artworks, dynamically guiding pre-trained stable diffusion to generate diverse and highly realistic artistic stylized images. DSPA can also generate random artistic image samples with the learned style information, providing a new idea for data augmentation. Besides, a Key Content Feature Prompt (KCFP) module is proposed to provide sufficient content prompts for pre-trained stable diffusion to preserve the detailed structure of the input content image. Extensive qualitative and quantitative experiments verify the effectiveness of our proposed method. Code is available: https://github.com/Jamie-Cheung/DyArtbank

cs.CV↗

arXiv2025

MC-VTON: Minimal Control Virtual Try-On Diffusion Transformer

Virtual try-on methods based on diffusion models achieve realistic try-on effects. They use an extra reference network or an additional image encoder to process multiple conditional image inputs, which adds complexity pre-processing and additional computational costs. Besides, they require more than 25 inference steps, bringing longer inference time. In this work, with the development of diffusion transformer (DiT), we rethink the necessity of additional reference network or image encoder and introduce MC-VTON, which leverages DiT's intrinsic backbone to seamlessly integrate minimal conditional try-on inputs. Compared to existing methods, the superiority of MC-VTON is demonstrated in four aspects: (1) Superior detail fidelity. Our DiT-based MC-VTON exhibits superior fidelity in preserving fine-grained details. (2) Simplified network and inputs. We remove any extra reference network or image encoder. We also remove unnecessary conditions like the long prompt, pose estimation, human parsing, and depth map. We require only the masked person image and the garment image. (3) Parameter-efficient training. To process the try-on task, we fine-tune the FLUX.1-dev with only 39.7M additional parameters (0.33% of the backbone parameters). (4) Less inference steps. We apply distillation diffusion on MC-VTON and only need 8 steps to generate a realistic try-on image, with only 86.8M additional parameters (0.72% of the backbone parameters). Experiments show that MC-VTON achieves superior qualitative and quantitative results with fewer condition inputs, trainable parameters, and inference steps than baseline methods.

cs.CV↗

arXiv2024

Tunable collective electromagnetic induced transparency-like effect due to coupling of dual-band bound states in the continuum

The coupling between dual-band or multi-band quasi-bound states in the continuum (q-BICs) is of great interest for their rich physics and promising applications. Here, we report tunable collective electromagnetic induced transparency-like (EIT-like) phenomenon due to coupling between dual-band collective electric dipolar and magnetic quadrupolar q-BICs, which are supported by an all-dielectric metasurface composed of periodic tilted silicon quadrumers. We show that this collective EIT-like phenomenon with strong slow light effect can be realized by varying the nanodisk diameter or the tilt angle, and that the transparency window wavelength, the quality factor, and the group index can all be tuned by changing the nanodisk size. We further find that as the nanodisk size decreases, the slow light effect becomes stronger, and higher sensitivity can be obtained for the refractive index sensing. Interestingly, the sensitivity first increases exponentially and then reaches a plateau as the nanodisk size decreases, or equivalently as the group index increases. We therefore expect this work will advance the understanding of the collective EIT-like effect due to coupling between q-BICs, and the findings will have potential applications in slow-light enhanced biochemical sensing.

physics.optics↗

arXiv2024

Significant loss suppression and large induced chirality via cooperative near- and far-field coupling in plasmonic dimer nanoantennas

Plasmonic nanoantennas containing nano-gaps support "hotspots" for greatly enhanced light-matter interactions, but suffer from inherent high losses, a long-standing issue that hinders practical applications. Here we report a strategy to significantly suppress the losses of plasmonic dimer nanoantennas. Specifically, by introducing the concept of cooperative near- and far-field coupling, we observed an unprecedented transition from the weak coupling of localized resonances to strong coupling of collective (nonlocal) resonances, showing robustness to the gap distance between the dimer. We develop a generalized lattice sum approximation model to describe this transition and reveal its origins: the off-diagonal element of the anisotropic polarizability tensor due to near-field coupling, and the anisotropic lattice sums due to far-field coupling. This strong coupling leads to loss-suppressed plasmonic resonances with large modulation depths and meanwhile extremely high measured quality factors up to 3120 in the near-infrared regime, exceeding the record in the near infrared regime. Additionally, high-$Q$ and large chiroptical responses can also be induced for achiral planar dimers under the critical coupling condition. This work paves an avenue toward extremely low-loss plasmonic devices, either chiral or not, for diverse important applications.

physics.optics↗

arXiv2024

Rethinking Video Deblurring with Wavelet-Aware Dynamic Transformer and Diffusion Model

Current video deblurring methods have limitations in recovering high-frequency information since the regression losses are conservative with high-frequency details. Since Diffusion Models (DMs) have strong capabilities in generating high-frequency details, we consider introducing DMs into the video deblurring task. However, we found that directly applying DMs to the video deblurring task has the following problems: (1) DMs require many iteration steps to generate videos from Gaussian noise, which consumes many computational resources. (2) DMs are easily misled by the blurry artifacts in the video, resulting in irrational content and distortion of the deblurred video. To address the above issues, we propose a novel video deblurring framework VD-Diff that integrates the diffusion model into the Wavelet-Aware Dynamic Transformer (WADT). Specifically, we perform the diffusion model in a highly compact latent space to generate prior features containing high-frequency information that conforms to the ground truth distribution. We design the WADT to preserve and recover the low-frequency information in the video while utilizing the high-frequency information generated by the diffusion model. Extensive experiments show that our proposed VD-Diff outperforms SOTA methods on GoPro, DVD, BSD, and Real-World Video datasets.

cs.CV↗

arXiv2024

Towards Highly Realistic Artistic Style Transfer via Stable Diffusion with Step-aware and Layer-aware Prompt

Artistic style transfer aims to transfer the learned artistic style onto an arbitrary content image, generating artistic stylized images. Existing generative adversarial network-based methods fail to generate highly realistic stylized images and always introduce obvious artifacts and disharmonious patterns. Recently, large-scale pre-trained diffusion models opened up a new way for generating highly realistic artistic stylized images. However, diffusion model-based methods generally fail to preserve the content structure of input content images well, introducing some undesired content structure and style patterns. To address the above problems, we propose a novel pre-trained diffusion-based artistic style transfer method, called LSAST, which can generate highly realistic artistic stylized images while preserving the content structure of input content images well, without bringing obvious artifacts and disharmonious style patterns. Specifically, we introduce a Step-aware and Layer-aware Prompt Space, a set of learnable prompts, which can learn the style information from the collection of artworks and dynamically adjusts the input images' content structure and style pattern. To train our prompt space, we propose a novel inversion method, called Step-ware and Layer-aware Prompt Inversion, which allows the prompt space to learn the style information of the artworks collection. In addition, we inject a pre-trained conditional branch of ControlNet into our LSAST, which further improved our framework's ability to maintain content structure. Extensive experiments demonstrate that our proposed method can generate more highly realistic artistic stylized images than the state-of-the-art artistic style transfer methods.

cs.CV↗