arXiv subjects

Shanshan Zhao

Publications and source records attributed to Shanshan Zhao.

At least 19 recordsLinked to original sources

arXiv2026

Test-Time Scaling for Video Diffusion Models via Diagnosis-Guided Candidate Recycling

Recent video diffusion models have achieved remarkable generation quality, but high-fidelity results still largely depend on closed-source systems or costly large-scale infrastructure. Test-time scaling (TTS) offers a training-free way to improve lightweight generators by spending additional inference compute, yet existing methods mostly remain within a noise-search paradigm: they sample, select, or perturb denoising trajectories and discard low-scoring candidates after expensive generation. This generate-and-discard process wastes not only computation but also the partial motion, layout, or appearance structure already encoded in recoverable samples. We present \textbf{GEARS} (\textbf{G}uided \textbf{E}diting for \textbf{A}daptive \textbf{R}ecycling \textbf{S}earch), a training-free framework that introduces {diagnosis-guided candidate recycling} into video TTS by turning such candidates into editable priors through a generation-evaluation-editing loop. GEARS consists of two collaborative components. The \textbf{Stage-Aware Scheduler} determines what to repair, when to repair it, and which candidates should be preserved, recycled, or discarded. The \textbf{Candidate Recycler} diagnoses recoverable failures from keyframes and multi-dimensional reward feedback, derives candidate-specific repair prompts, and repairs the corresponding candidates through manifold-aware latent SDEdit. The repaired candidates are recycled into the search pool, creating refinement paths beyond standard noise perturbation while preserving useful structure. Under matched NFE budgets, GEARS consistently outperforms existing video TTS methods on VBench, bringing a 1.3B model to a total score comparable to a 14B counterpart, and ablations verify the necessity of adaptive scheduling, diagnosis-conditioned editing, and manifold-aware re-denoising. Code is available on GitHub.

cs.CV↗

arXiv2026

D2-ScaleAgent: Dual-Dimensional Scaling for Long Document Understanding

Multi-modal retrieval-augmented generation (RAG) is a key technique for visually rich long document understanding. Existing multi-modal RAG methods are progressively advancing toward multi-agent systems: they first retrieve relevant pages based on a query, and then iteratively understand information within those pages. However, these methods typically rely on fixed workflows and lack the ability to dynamically scale computation at test time, often leading to insufficient evidence. To address this, we propose D2-ScaleAgent, an agentic framework that introduces a dual-dimensional scaling paradigm for retrieval and reasoning. The core of D2-ScaleAgent is a Verifier agent-driven dynamic routing loop based on the intrinsic difficulty of the query, centered around a continuously updated evidence bank that serves as the agent's dynamic working memory: when retrieval needs to be expanded, the agent routes outward (retrieval scaling), decomposing the query into attributes and performing parallel page retrieval, followed by adaptive pruning to ensure comprehensive evidence coverage. When fine-grained reasoning is required, the agent routes inward (reasoning scaling), dynamically selecting sub-agents with varying granularity and count to extract evidence from pages. Finally, D2-ScaleAgent achieves logical closure over the evidence chain. Extensive experiments demonstrate that D2-ScaleAgent is effective on long and visually rich document benchmarks like MMLongBench-Doc, LongDocURL, etc.

cs.CL↗

arXiv2026

FreeSpec: Training-Free Long Video Generation via Singular-Spectrum Reconstruction

Video diffusion models perform well in short-video synthesis, but their training-free extension to long videos often suffers from content drift, temporal inconsistency, and over-smoothed dynamics. Existing methods improve temporal consistency by combining a global branch with a local branch, but they often further decompose appearance consistency and temporal dynamics within each branch using predefined criteria. This assignment is unreliable when appearance and action progression are tightly coupled, such as in camera motion and sequential motion. We analyze the video temporal extension issue from a singular-spectrum perspective and show that enlarged self-attention windows induce spectral concentration: spectral energy becomes dominated by a few low-rank singular directions, preserving coarse structure but suppressing high-rank spatial details and motion-rich temporal variations. To mitigate this problem, we propose FreeSpec, a training-free spectral reconstruction framework for long-video generation. FreeSpec decomposes global and local features with singular value decomposition, and uses the global branch as low-rank spectral guidance and the local branch as a high-rank reconstruction basis. This spectrum-level fusion avoids the rigid feature partitioning of previous decomposition rules, preserving long-range consistency while better retaining spatial details and temporal dynamics. Experiments on Wan2.1 and LTX-Video demonstrate that FreeSpec improves long-video generation, especially for temporal dynamics, while maintaining strong visual quality and temporal consistency. Project demo: https://fdchen24.github.io/FreeSpec-Website/.

cs.CV↗

arXiv2026

InpaintHuman: Reconstructing Occluded Humans with Multi-Scale UV Mapping and Identity-Preserving Diffusion Inpainting

Reconstructing complete and animatable 3D human avatars from monocular videos remains challenging, particularly under severe occlusions. While 3D Gaussian Splatting has enabled photorealistic human rendering, existing methods struggle with incomplete observations, often producing corrupted geometry and temporal inconsistencies. We present InpaintHuman, a novel method for generating high-fidelity, complete, and animatable avatars from occluded monocular videos. Our approach introduces two key innovations: (i) a multi-scale UV-parameterized representation with hierarchical coarse-to-fine feature interpolation, enabling robust reconstruction of occluded regions while preserving geometric details; and (ii) an identity-preserving diffusion inpainting module that integrates textual inversion with semantic-conditioned guidance for subject-specific, temporally coherent completion. Unlike SDS-based methods, our approach employs direct pixel-level supervision to ensure identity fidelity. Experiments on synthetic benchmarks (PeopleSnapshot, ZJU-MoCap) and real-world scenarios (OcMotion) demonstrate competitive performance with consistent improvements in reconstruction quality across diverse poses and viewpoints.

cs.CV↗

arXiv2025

Omni-View: Unlocking How Generation Facilitates Understanding in Unified 3D Model based on Multiview images

This paper presents Omni-View, which extends the unified multimodal understanding and generation to 3D scenes based on multiview images, exploring the principle that "generation facilitates understanding". Consisting of understanding model, texture module, and geometry module, Omni-View jointly models scene understanding, novel view synthesis, and geometry estimation, enabling synergistic interaction between 3D scene understanding and generation tasks. By design, it leverages the spatiotemporal modeling capabilities of its texture module responsible for appearance synthesis, alongside the explicit geometric constraints provided by its dedicated geometry module, thereby enriching the model's holistic understanding of 3D scenes. Trained with a two-stage strategy, Omni-View achieves a state-of-the-art score of 55.4 on the VSI-Bench benchmark, outperforming existing specialized 3D understanding models, while simultaneously delivering strong performance in both novel view synthesis and 3D scene generation. The code and pretraiend models are open-sourced at https://github.com/AIDC-AI/Omni-View.

cs.CV↗

arXiv2025

Scaling Beyond Context: A Survey of Multimodal Retrieval-Augmented Generation for Document Understanding

Document understanding is critical for applications from financial analysis to scientific discovery. Current approaches, whether OCR-based pipelines feeding Large Language Models (LLMs) or native Multimodal LLMs (MLLMs), face key limitations: the former loses structural detail, while the latter struggles with context modeling. Retrieval-Augmented Generation (RAG) helps ground models in external data, but documents' multimodal nature, i.e., combining text, tables, charts, and layout, demands a more advanced paradigm: Multimodal RAG. This approach enables holistic retrieval and reasoning across all modalities, unlocking comprehensive document intelligence. Recognizing its importance, this paper presents a systematic survey of Multimodal RAG for document understanding. We propose a taxonomy based on domain, retrieval modality, and granularity, and review advances involving graph structures and agentic frameworks. We also summarize key datasets, benchmarks, applications and industry deployment, and highlight open challenges in efficiency, fine-grained representation, and robustness, providing a roadmap for future progress in document AI.

cs.CL↗

arXiv2025

Cross3DReg: Towards a Large-scale Real-world Cross-source Point Cloud Registration Benchmark

Cross-source point cloud registration, which aims to align point cloud data from different sensors, is a fundamental task in 3D vision. However, compared to the same-source point cloud registration, cross-source registration faces two core challenges: the lack of publicly available large-scale real-world datasets for training the deep registration models, and the inherent differences in point clouds captured by multiple sensors. The diverse patterns induced by the sensors pose great challenges in robust and accurate point cloud feature extraction and matching, which negatively influence the registration accuracy. To advance research in this field, we construct Cross3DReg, the currently largest and real-world multi-modal cross-source point cloud registration dataset, which is collected by a rotating mechanical lidar and a hybrid semi-solid-state lidar, respectively. Moreover, we design an overlap-based cross-source registration framework, which utilizes unaligned images to predict the overlapping region between source and target point clouds, effectively filtering out redundant points in the irrelevant regions and significantly mitigating the interference caused by noise in non-overlapping areas. Then, a visual-geometric attention guided matching module is proposed to enhance the consistency of cross-source point cloud features by fusing image and geometric information to establish reliable correspondences and ultimately achieve accurate and robust registration. Extensive experiments show that our method achieves state-of-the-art registration performance. Our framework reduces the relative rotation error (RRE) and relative translation error (RTE) by $63.2\%$ and $40.2\%$, respectively, and improves the registration recall (RR) by $5.4\%$, which validates its effectiveness in achieving accurate cross-source registration.

cs.CV↗

arXiv2025

Ovis2.5 Technical Report

We present Ovis2.5, a successor to Ovis2 designed for native-resolution visual perception and strong multimodal reasoning. Ovis2.5 integrates a native-resolution vision transformer that processes images at their native, variable resolutions, avoiding the degradation from fixed-resolution tiling and preserving both fine detail and global layout -- crucial for visually dense content like complex charts. To strengthen reasoning, we train the model to move beyond linear chain-of-thought and perform reflection -- including self-checking and revision. This advanced capability is exposed as an optional "thinking mode" at inference time, allowing users to trade latency for enhanced accuracy on difficult inputs. The model is trained via a comprehensive five-phase curriculum that progressively builds its skills. The process begins with foundational visual and multimodal pretraining, advances through large-scale instruction tuning, and culminates in alignment and reasoning enhancement using DPO and GRPO. To scale these upgrades efficiently, we employ multimodal data packing and hybrid parallelism, yielding a significant end-to-end speedup. We release two open-source models: Ovis2.5-9B and Ovis2.5-2B. The latter continues the "small model, big performance" philosophy of Ovis2, making it ideal for resource-constrained, on-device scenarios. On the OpenCompass multimodal leaderboard, Ovis2.5-9B averages 78.3, marking a substantial improvement over its predecessor, Ovis2-8B, and achieving state-of-the-art results among open-source MLLMs in the sub-40B parameter range; Ovis2.5-2B scores 73.9, establishing SOTA for its size. Beyond aggregate scores, Ovis2.5 achieves leading results on STEM benchmarks, exhibits strong capabilities on grounding and video tasks, and achieves open-source SOTA at its scale for complex chart analysis.

cs.CV↗

arXiv2025

Lunar Orbital VLBI Experiment: motivation, scientific purposes and status

The Lunar Orbital VLBI Experiment (LOVEX) is a scientific component of the Chinese Lunar Exploration Project (CLEP) Chang'E-7. The spaceborne component of LOVEX is implemented onboard the relay satellite QueQiao-2, which was launched on 2024 March 20, and later placed into an elliptical selenocentric orbit. The LOVEX-specific payload consists of an X-band cryogenic receiver, a hydrogen maser frequency standard, and VLBI data formatting and acquisition electronics. Several components of the QueQiao-2 nominal onboard instrumentation, such as the 4.2-meter antenna, the data storage device, and the downlink communication system, contribute to the overall spaceborne VLBI instrumentation. This allows us to form a space radio telescope capable of co-observing with Earth-based radio telescopes in VLBI mode. In this space VLBI system, the length of the baseline extends up to approximately 380,000 km. This paper presents the LOVEX scientific objectives, architecture, instrumentation, pre-launch tests, in-flight verification and calibration, and the first in-flight detections of interferometric response (''fringes'') achieved through observations of the quasar AO 0235+164 and the Chang'E-6 orbital module, positioned at the Sun-Earth Lagrange point L2. These initial results demonstrate the successful performance of LOVEX, verifying its capability for both astronomical and spacecraft tracking observations at ultra-long VLBI baselines.

astro-ph.IM↗

arXiv2025

Harnessing Text-to-Image Diffusion Models for Point Cloud Self-Supervised Learning

Diffusion-based models, widely used in text-to-image generation, have proven effective in 2D representation learning. Recently, this framework has been extended to 3D self-supervised learning by constructing a conditional point generator for enhancing 3D representations. However, its performance remains constrained by the 3D diffusion model, which is trained on the available 3D datasets with limited size. We hypothesize that the robust capabilities of text-to-image diffusion models, particularly Stable Diffusion (SD), which is trained on large-scale datasets, can help overcome these limitations. To investigate this hypothesis, we propose PointSD, a framework that leverages the SD model for 3D self-supervised learning. By replacing the SD model's text encoder with a 3D encoder, we train a point-to-image diffusion model that allows point clouds to guide the denoising of rendered noisy images. With the trained point-to-image diffusion model, we use noise-free images as the input and point clouds as the condition to extract SD features. Next, we train a 3D backbone by aligning its features with these SD features, thereby facilitating direct semantic learning. Comprehensive experiments on downstream point cloud tasks and ablation studies demonstrate that the SD model can enhance point cloud self-supervised learning. Code is publicly available at https://github.com/wdttt/PointSD.

cs.CV↗

arXiv2025

Ovis-U1 Technical Report

In this report, we introduce Ovis-U1, a 3-billion-parameter unified model that integrates multimodal understanding, text-to-image generation, and image editing capabilities. Building on the foundation of the Ovis series, Ovis-U1 incorporates a diffusion-based visual decoder paired with a bidirectional token refiner, enabling image generation tasks comparable to leading models like GPT-4o. Unlike some previous models that use a frozen MLLM for generation tasks, Ovis-U1 utilizes a new unified training approach starting from a language model. Compared to training solely on understanding or generation tasks, unified training yields better performance, demonstrating the enhancement achieved by integrating these two tasks. Ovis-U1 achieves a score of 69.6 on the OpenCompass Multi-modal Academic Benchmark, surpassing recent state-of-the-art models such as Ristretto-3B and SAIL-VL-1.5-2B. In text-to-image generation, it excels with scores of 83.72 and 0.89 on the DPG-Bench and GenEval benchmarks, respectively. For image editing, it achieves 4.00 and 6.42 on the ImgEdit-Bench and GEdit-Bench-EN, respectively. As the initial version of the Ovis unified model series, Ovis-U1 pushes the boundaries of multimodal understanding, generation, and editing.

cs.CV↗

arXiv2025

High-quality Pseudo-labeling for Point Cloud Segmentation with Scene-level Annotation

This paper investigates indoor point cloud semantic segmentation under scene-level annotation, which is less explored compared to methods relying on sparse point-level labels. In the absence of precise point-level labels, current methods first generate point-level pseudo-labels, which are then used to train segmentation models. However, generating accurate pseudo-labels for each point solely based on scene-level annotations poses a considerable challenge, substantially affecting segmentation performance. Consequently, to enhance accuracy, this paper proposes a high-quality pseudo-label generation framework by exploring contemporary multi-modal information and region-point semantic consistency. Specifically, with a cross-modal feature guidance module, our method utilizes 2D-3D correspondences to align point cloud features with corresponding 2D image pixels, thereby assisting point cloud feature learning. To further alleviate the challenge presented by the scene-level annotation, we introduce a region-point semantic consistency module. It produces regional semantics through a region-voting strategy derived from point-level semantics, which are subsequently employed to guide the point-level semantic predictions. Leveraging the aforementioned modules, our method can rectify inaccurate point-level semantic predictions during training and obtain high-quality pseudo-labels. Significant improvements over previous works on ScanNet v2 and S3DIS datasets under scene-level annotation can demonstrate the effectiveness. Additionally, comprehensive ablation studies validate the contributions of our approach's individual components. The code is available at https://github.com/LHDuan/WSegPC .

cs.CV↗

arXiv2025

Strategic A/B testing via Maximum Probability-driven Two-armed Bandit

Detecting a minor average treatment effect is a major challenge in large-scale applications, where even minimal improvements can have a significant economic impact. Traditional methods, reliant on normal distribution-based or expanded statistics, often fail to identify such minor effects because of their inability to handle small discrepancies with sufficient sensitivity. This work leverages a counterfactual outcome framework and proposes a maximum probability-driven two-armed bandit (TAB) process by weighting the mean volatility statistic, which controls Type I error. The implementation of permutation methods further enhances the robustness and efficacy. The established strategic central limit theorem (SCLT) demonstrates that our approach yields a more concentrated distribution under the null hypothesis and a less concentrated one under the alternative hypothesis, greatly improving statistical power. The experimental results indicate a significant improvement in the A/B testing, highlighting the potential to reduce experimental costs while maintaining high statistical power.

stat.ML↗

arXiv2025

Active RIS Enabled NLoS LEO Satellite Communications: A Three-timescale Optimization Framework

In this letter, we study an active reconfigurable intelligent surfaces (RIS) assisted Low Earth orbit (LEO) satellite communications under non-line-of-sight (NLoS) scenarios, where the active RIS is deployed to create visual line-of-sight links for reliable communication. To address the challenges of high energy consumption caused by frequent beamforming updates in active RIS, we propose a three-timescale optimization framework that jointly designs the transmit beamforming, RIS beamforming, and RIS direction vectors based on their characteristics. The goal is to maximize the system achievable rate while reducing energy consumption by controlling the RIS beamforming switching frequency. Then, a two-layer solution framework is developed, incorporating fractional programming (FP), alternating optimization (AO), successive approximation (SCA), and penalty-based methods, to obtain the optimized solution. Simulation results demonstrate that the proposed scheme can effectively improve system performance and reduce the energy consumption of the active RIS.

eess.SP↗

arXiv2025

Unified Multimodal Understanding and Generation Models: Advances, Challenges, and Opportunities

Recent years have seen remarkable progress in both multimodal understanding models and image generation models. Despite their respective successes, these two domains have evolved independently, leading to distinct architectural paradigms: While autoregressive-based architectures have dominated multimodal understanding, diffusion-based models have become the cornerstone of image generation. Recently, there has been growing interest in developing unified frameworks that integrate these tasks. The emergence of GPT-4o's new capabilities exemplifies this trend, highlighting the potential for unification. However, the architectural differences between the two domains pose significant challenges. To provide a clear overview of current efforts toward unification, we present a comprehensive survey aimed at guiding future research. First, we introduce the foundational concepts and recent advancements in multimodal understanding and text-to-image generation models. Next, we review existing unified models, categorizing them into three main architectural paradigms: diffusion-based, autoregressive-based, and hybrid approaches that fuse autoregressive and diffusion mechanisms. For each category, we analyze the structural designs and innovations introduced by related works. Additionally, we compile datasets and benchmarks tailored for unified models, offering resources for future exploration. Finally, we discuss the key challenges facing this nascent field, including tokenization strategy, cross-modal attention, and data. As this area is still in its early stages, we anticipate rapid advancements and will regularly update this survey. Our goal is to inspire further research and provide a valuable reference for the community. The references associated with this survey are available on GitHub (https://github.com/AIDC-AI/Awesome-Unified-Multimodal-Models).

cs.CV↗

arXiv2025

JointTuner: Appearance-Motion Adaptive Joint Training for Customized Video Generation

Recent advancements in customized video generation have led to significant improvements in the simultaneous adaptation of appearance and motion. Typically, decoupling the appearance and motion training, prior methods often introduce concept interference, resulting in inaccurate rendering of appearance features or motion patterns. In addition, these methods often suffer from appearance contamination, in which background and foreground elements from reference videos distort the customized video. This paper aims to alleviate these issues by proposing JointTuner. The core motivation of our JointTuner is to enable joint optimization of both appearance and motion components, upon which two key innovations are developed, i.e., Gated Low-Rank Adaptation (GLoRA) and Appearance-independent Temporal Loss (AiT Loss). Specifically, GLoRA uses a context-aware activation layer, analogous to a gating regulator, to dynamically steer LoRA modules toward learning either appearance or motion while maintaining spatio-temporal consistency. Moreover, with the finding that channel-temporal shift noise suppresses appearance-related low-frequencies while enhancing motion-related high-frequencies, we designed the AiT Loss. This loss adds the same shift to the diffusion model's predicted noise during fine-tuning, forcing the model to prioritize learning motion patterns. JointTuner's architecture-agnostic design supports both UNet (e.g., ZeroScope) and Diffusion Transformer (e.g., CogVideoX) backbones, ensuring its customization capabilities scale with the evolution of foundational video models. Furthermore, we present a systematic evaluation framework for appearance-motion combined customization, covering 90 combinations evaluated along four critical dimensions: semantic alignment, motion dynamism, temporal consistency, and perceptual quality. Our project homepage is available online.

cs.CV↗

arXiv2024

UNIC-Adapter: Unified Image-instruction Adapter with Multi-modal Transformer for Image Generation

Recently, text-to-image generation models have achieved remarkable advancements, particularly with diffusion models facilitating high-quality image synthesis from textual descriptions. However, these models often struggle with achieving precise control over pixel-level layouts, object appearances, and global styles when using text prompts alone. To mitigate this issue, previous works introduce conditional images as auxiliary inputs for image generation, enhancing control but typically necessitating specialized models tailored to different types of reference inputs. In this paper, we explore a new approach to unify controllable generation within a single framework. Specifically, we propose the unified image-instruction adapter (UNIC-Adapter) built on the Multi-Modal-Diffusion Transformer architecture, to enable flexible and controllable generation across diverse conditions without the need for multiple specialized models. Our UNIC-Adapter effectively extracts multi-modal instruction information by incorporating both conditional images and task instructions, injecting this information into the image generation process through a cross-attention mechanism enhanced by Rotary Position Embedding. Experimental results across a variety of tasks, including pixel-level spatial control, subject-driven image generation, and style-image-based image synthesis, demonstrate the effectiveness of our UNIC-Adapter in unified controllable image generation.

cs.CV↗

arXiv2024

When ControlNet Meets Inexplicit Masks: A Case Study of ControlNet on its Contour-following Ability

ControlNet excels at creating content that closely matches precise contours in user-provided masks. However, when these masks contain noise, as a frequent occurrence with non-expert users, the output would include unwanted artifacts. This paper first highlights the crucial role of controlling the impact of these inexplicit masks with diverse deterioration levels through in-depth analysis. Subsequently, to enhance controllability with inexplicit masks, an advanced Shape-aware ControlNet consisting of a deterioration estimator and a shape-prior modulation block is devised. The deterioration estimator assesses the deterioration factor of the provided masks. Then this factor is utilized in the modulation block to adaptively modulate the model's contour-following ability, which helps it dismiss the noise part in the inexplicit masks. Extensive experiments prove its effectiveness in encouraging ControlNet to interpret inaccurate spatial conditions robustly rather than blindly following the given contours, suitable for diverse kinds of conditions. We showcase application scenarios like modifying shape priors and composable shape-controllable generation. Codes are available at github.

cs.CV↗