Searcharxiv⌕ Search

arXiv subjects

Xinyuan Chen

Publications and source records attributed to Xinyuan Chen.

At least 37 records · Page 2Linked to original sources
arXiv2026

XTransfer: Modality-Agnostic Few-Shot Model Transfer for Human Sensing at the Edge

Deep learning for human sensing on edge systems presents significant potential for smart applications. However, its training and development are hindered by the limited availability of sensor data and resource constraints of edge systems. While transferring pre-trained models to different sensing applications is promising, existing methods often require extensive sensor data and computational resources, resulting in high costs and limited transferability. In this paper, we propose XTransfer, a first-of-its-kind method enabling modality-agnostic, few-shot model transfer with resource-efficient design. XTransfer flexibly uses pre-trained models and transfers knowledge across different modalities by (i) model repairing that safely mitigates modality shift by adapting pre-trained layers with only few sensor data, and (ii) layer recombining that efficiently searches and recombines layers of interest from source models in a layer-wise manner to restructure models. We benchmark various baselines across diverse human sensing datasets spanning different modalities. The results show that XTransfer achieves state-of-the-art performance while significantly reducing the costs of sensor data collection, model training, and edge deployment.

cs.CV↗
arXiv2026

PARE: Pruning and Adaptive Routing for Efficient Video Generation

Video Diffusion Transformers (DiTs) generate high-quality videos but demand substantial compute due to wide blocks, deep architectures, and iterative sampling. Recent methods reduce cost by compressing width, depth, or sampling steps, but typically commit to a fixed architecture that cannot adapt to individual inputs or denoising stages. We propose PARE (Pruning and Adaptive Routing for Efficient video generation), which jointly compresses width and depth with structure-aware pruning and input-adaptive routing. For width, we observe that attention heads specialize into spatial and temporal roles, and design importance scoring that accounts for this distinction to prevent motion-critical temporal heads from being pruned prematurely. For depth, we train a lightweight router conditioned on denoising timestep and visual content to dynamically select which blocks to execute at each step, enabling per-input compute adaptation rather than static block removal. A progressive pipeline first recovers width-pruned quality via distillation, then jointly optimizes the student and router to decouple the two learning objectives. Experiments on Wan2.1-14B for both image-to-video and text-to-video generation show that PARE substantially reduces per-step computation while preserving quality across VBench dimensions, and composes with step distillation for further acceleration.

cs.CV↗
arXiv2026

RAPO++: Cross-Stage Prompt Optimization for Text-to-Video Generation via Data Alignment and Test-Time Scaling

Prompt design plays a crucial role in text-to-video (T2V) generation, yet user-provided prompts are often short, unstructured, and misaligned with training data, limiting the generative potential of diffusion-based T2V models. We present \textbf{RAPO++}, a cross-stage prompt optimization framework that unifies training-data--aligned refinement, test-time iterative scaling, and large language model (LLM) fine-tuning to substantially improve T2V generation without modifying the underlying generative backbone. In \textbf{Stage 1}, Retrieval-Augmented Prompt Optimization (RAPO) enriches user prompts with semantically relevant modifiers retrieved from a relation graph and refactors them to match training distributions, enhancing compositionality and multi-object fidelity. \textbf{Stage 2} introduces Sample-Specific Prompt Optimization (SSPO), a closed-loop mechanism that iteratively refines prompts using multi-source feedback -- including semantic alignment, spatial fidelity, temporal coherence, and task-specific signals such as optical flow -- yielding progressively improved video generation quality. \textbf{Stage 3} leverages optimized prompt pairs from SSPO to fine-tune the rewriter LLM, internalizing task-specific optimization patterns and enabling efficient, high-quality prompt generation even before inference. Extensive experiments across five state-of-the-art T2V models and five benchmarks demonstrate that RAPO++ achieves significant gains in semantic alignment, compositional reasoning, temporal stability, and physical plausibility, outperforming existing methods by large margins. Our results highlight RAPO++ as a model-agnostic, cost-efficient, and scalable solution that sets a new standard for prompt optimization in T2V generation. The code is available at https://github.com/Vchitect/RAPO.

cs.CV↗
arXiv2026

Asymptotic theory of rerandomization for survival analysis

Rerandomization systematically reduces chance imbalance and can improve the efficiency of the average treatment effect estimator in randomized experiments. While the asymptotic properties of finite-dimensional M-estimators under rerandomization have been established, existing theory does not directly address survival outcomes under censoring, where the target estimand involves infinite-dimensional functional parameters. This article establishes the uniform weak convergence of treatment-specific survival function estimators under rerandomization and stratified rerandomization. We prove that the Kaplan-Meier and inverse probability of censoring weighted Kaplan-Meier estimators converge to tight limiting processes with reduced pointwise asymptotic variances. Furthermore, we prove that the pointwise asymptotic variance of the debiased machine learning survival function estimator remains invariant under rerandomization, a consequence of the Neyman orthogonality. Simulations and a real data example are used to illustrate the theoretical results. Our results characterize the geometric interplay between restricted randomization designs and analysis-stage covariate adjustment for functional target estimands in survival analysis.

stat.ME↗
arXiv2026

Uni-Animator: Towards Unified Visual Colorization

We propose Uni-Animator, a novel Diffusion Transformer (DiT)-based framework for unified image and video sketch colorization. Existing sketch colorization methods struggle to unify image and video tasks, suffering from imprecise color transfer with single or multiple references, inadequate preservation of high-frequency physical details, and compromised temporal coherence with motion artifacts in large-motion scenes. To tackle imprecise color transfer, we introduce visual reference enhancement via instance patch embedding, enabling precise alignment and fusion of reference color information. To resolve insufficient physical detail preservation, we design physical detail reinforcement using physical features that effectively capture and retain high-frequency textures. To mitigate motion-induced temporal inconsistency, we propose sketch-based dynamic RoPE encoding that adaptively models motion-aware spatial-temporal dependencies. Extensive experimental results demonstrate that Uni-Animator achieves competitive performance on both image and video sketch colorization, matching that of task-specific methods while unlocking unified cross-domain capabilities with high detail fidelity and robust temporal consistency.

cs.CV↗
arXiv2026

CineTrans: Learning to Generate Videos with Cinematic Transitions via Masked Diffusion Models

Despite significant advances in video synthesis, research into multi-shot video generation remains in its infancy. Even with scaled-up models and massive datasets, the shot transition capabilities remain rudimentary and unstable, largely confining generated videos to single-shot sequences. In this work, we introduce CineTrans, a novel framework for generating coherent multi-shot videos with cinematic, film-style transitions. To facilitate insights into the film editing style, we construct a multi-shot video-text dataset Cine250K with detailed shot annotations. Furthermore, our analysis of existing video diffusion models uncovers a correspondence between attention maps in the diffusion model and shot boundaries, which we leverage to design a mask-based control mechanism that enables transitions at arbitrary positions and transfers effectively in a training-free setting. After fine-tuning on our dataset with the mask mechanism, CineTrans produces cinematic multi-shot sequences while adhering to the film editing style, avoiding unstable transitions or naive concatenations. Finally, we propose specialized evaluation metrics for transition control, temporal consistency and overall quality, and demonstrate through extensive experiments that CineTrans significantly outperforms existing baselines across all criteria.

cs.CV↗
arXiv2026

Model-assisted inference for dynamic causal effects in staggered rollout cluster randomized experiments

Staggered rollout cluster randomized experiments (SR-CREs) involve sequential treatment adoption across clusters, requiring analysis methods that address a general class of dynamic causal effects, anticipation, and non-ignorable cluster-period sizes. Without imposing any outcome modeling assumptions, we study regression estimators using individual data, cluster-period averages, and scaled cluster-period totals, with and without covariate adjustment from a design-based perspective. We establish consistency and asymptotic normality of each estimator under a randomization-based framework and prove that the associated variance estimators are asymptotically conservative in the Löwner ordering. Furthermore, we conduct a unified efficiency comparison of the estimators and provide recommendations. We highlight the efficiency advantage of using estimators based on scaled cluster-period totals with covariate adjustment over their counterparts using individual-level data and cluster-period averages. Our results rigorously justify linear regression estimators as model-assisted methods to address an entire class of dynamic causal effects in SR-CREs.

stat.ME↗
arXiv2026

Optimal estimation of generalized causal effects in cluster-randomized trials with multiple outcomes

Cluster-randomized trials (CRTs) are widely used to evaluate group-level interventions and increasingly collect multiple outcomes capturing complementary dimensions of benefit and risk. Investigators often seek a single global summary of treatment effect, yet existing methods largely focus on single-outcome estimands or rely on model-based procedures with unclear causal interpretation or limited robustness. We develop a unified potential outcomes framework for generalized treatment effects with multiple outcomes in CRTs, accommodating both non-prioritized and prioritized outcome settings. The proposed cluster-pair and individual-pair causal estimands are defined through flexible pairwise contrast functions and explicitly account for potentially informative cluster sizes. We establish nonparametric estimation via weighted clustered U-statistics and derive efficient influence functions to construct covariate-adjusted estimators that integrate debiased machine learning with U-statistics. The resulting estimators are consistent and asymptotically normal, attain the semiparametric efficiency bounds under mild regularity conditions, and have analytically tractable variance estimators that are proven to be consistent under cross-fitting. Simulations and an application to a CRT for chronic pain management illustrate the practical utility of the proposed methods.

stat.ME↗
arXiv2025

Principal stratification with U-statistics under principal ignorability

Principal stratification is a popular framework for causal inference in the presence of an intermediate outcome. While the principal average treatment effects are the standard target of inference, they may be insufficient when interest lies in the relative ordering of potential outcomes within a principal stratum. We introduce the principal generalized causal effect estimands to accommodate nonlinear contrast functions, providing robust, probability-scale summaries suitable for ordinal outcomes and win-loss comparisons with composite endpoints. Under principal ignorability, we expand the theoretical results in Jiang et al. (2022, JRSSB) to a broader class of causal estimands in the presence of a binary intermediate variable. We develop nonparametric identification results and derive efficient influence functions for the generalized causal estimands in principal stratification analyses. These efficient influence functions motivate multiply robust estimators and lay the ground for obtaining efficient debiased machine learning estimators via cross-fitting based on U-statistics. The proposed methods are illustrated through simulations and the analysis of a data example.

stat.ME↗
arXiv2025

VDOT: Efficient Unified Video Creation via Optimal Transport Distillation

The rapid development of generative models has significantly advanced image and video applications. Among these, video creation, aimed at generating videos under various conditions, has gained substantial attention. However, existing video creation models either focus solely on a few specific conditions or suffer from excessively long generation times due to complex model inference, making them impractical for real-world applications. To mitigate these issues, we propose an efficient unified video creation model, named VDOT. Concretely, we model the training process with the distribution matching distillation (DMD) paradigm. Instead of using the Kullback-Leibler (KL) minimization, we additionally employ a novel computational optimal transport (OT) technique to optimize the discrepancy between the real and fake score distributions. The OT distance inherently imposes geometric constraints, mitigating potential zero-forcing or gradient collapse issues that may arise during KL-based distillation within the few-step generation scenario, and thus, enhances the efficiency and stability of the distillation process. Further, we integrate a discriminator to enable the model to perceive real video data, thereby enhancing the quality of generated videos. To support training unified video creation models, we propose a fully automated pipeline for video data annotation and filtering that accommodates multiple video creation tasks. Meanwhile, we curate a unified testing benchmark, UVCBench, to standardize evaluation. Experiments demonstrate that our 4-step VDOT outperforms or matches other baselines with 100 denoising steps.

cs.CV↗
arXiv2025

ShotDirector: Directorially Controllable Multi-Shot Video Generation with Cinematographic Transitions

Shot transitions play a pivotal role in multi-shot video generation, as they determine the overall narrative expression and the directorial design of visual storytelling. However, recent progress has primarily focused on low-level visual consistency across shots, neglecting how transitions are designed and how cinematographic language contributes to coherent narrative expression. This often leads to mere sequential shot changes without intentional film-editing patterns. To address this limitation, we propose ShotDirector, an efficient framework that integrates parameter-level camera control and hierarchical editing-pattern-aware prompting. Specifically, we adopt a camera control module that incorporates 6-DoF poses and intrinsic settings to enable precise camera information injection. In addition, a shot-aware mask mechanism is employed to introduce hierarchical prompts aware of professional editing patterns, allowing fine-grained control over shot content. Through this design, our framework effectively combines parameter-level conditions with high-level semantic guidance, achieving film-like controllable shot transitions. To facilitate training and evaluation, we construct ShotWeaver40K, a dataset that captures the priors of film-like editing patterns, and develop a set of evaluation metrics for controllable multi-shot video generation. Extensive experiments demonstrate the effectiveness of our framework.

cs.CV↗
arXiv2025

On Anticipation Effect in Stepped Wedge Cluster Randomized Trials

In stepped wedge cluster randomized trials (SW-CRTs), the intervention is rolled out to clusters over multiple periods. A standard approach for analyzing SW-CRTs utilizes the linear mixed model, where the treatment effect is only present after the treatment adoption, under the assumption of no anticipation. This assumption, however, may not always hold in practice because stakeholders, providers, or individuals who are aware of the treatment adoption timing (especially when blinding is challenging or infeasible) can inadvertently change their behaviors in anticipation of the forthcoming intervention. We provide an analytical framework to address the anticipation effect in SW-CRTs and study its impact. We derive expectations of the estimators based on a collection of linear mixed models and demonstrate that when the anticipation effect is ignored, these estimators give biased estimates of the treatment effect. We also provide updated sample size formulas that explicitly account for anticipation effects, exposure-time heterogeneity, or both in SW-CRTs and illustrate their impact on study power. Through simulation studies and empirical analyses, we compare the treatment effect estimators with and without adjusting for anticipation, and provide some practical considerations.

stat.ME↗
arXiv2025

LIA-X: Interpretable Latent Portrait Animator

We introduce LIA-X, a novel interpretable portrait animator designed to transfer facial dynamics from a driving video to a source portrait with fine-grained control. LIA-X is an autoencoder that models motion transfer as a linear navigation of motion codes in latent space. Crucially, it incorporates a novel Sparse Motion Dictionary that enables the model to disentangle facial dynamics into interpretable factors. Deviating from previous 'warp-render' approaches, the interpretability of the Sparse Motion Dictionary allows LIA-X to support a highly controllable 'edit-warp-render' strategy, enabling precise manipulation of fine-grained facial semantics in the source portrait. This helps to narrow initial differences with the driving video in terms of pose and expression. Moreover, we demonstrate the scalability of LIA-X by successfully training a large-scale model with approximately 1 billion parameters on extensive datasets. Experimental results show that our proposed method outperforms previous approaches in both self-reenactment and cross-reenactment tasks across several benchmarks. Additionally, the interpretable and controllable nature of LIA-X supports practical applications such as fine-grained, user-guided image and video editing, as well as 3D-aware portrait video manipulation.

cs.CV↗
arXiv2025

Consistent and Controllable Image Animation with Motion Linear Diffusion Transformers

Image animation has seen significant progress, driven by the powerful generative capabilities of diffusion models. However, maintaining appearance consistency with static input images and mitigating abrupt motion transitions in generated animations remain persistent challenges. While text-to-video (T2V) generation has demonstrated impressive performance with diffusion transformer models, the image animation field still largely relies on U-Net-based diffusion models, which lag behind the latest T2V approaches. Moreover, the quadratic complexity of vanilla self-attention mechanisms in Transformers imposes heavy computational demands, making image animation particularly resource-intensive. To address these issues, we propose MiraMo, a framework designed to enhance efficiency, appearance consistency, and motion smoothness in image animation. Specifically, MiraMo introduces three key elements: (1) A foundational text-to-video architecture replacing vanilla self-attention with efficient linear attention to reduce computational overhead while preserving generation quality; (2) A novel motion residual learning paradigm that focuses on modeling motion dynamics rather than directly predicting frames, improving temporal consistency; and (3) A DCT-based noise refinement strategy during inference to suppress sudden motion artifacts, complemented by a dynamics control module to balance motion smoothness and expressiveness. Extensive experiments against state-of-the-art methods validate the superiority of MiraMo in generating consistent, smooth, and controllable animations with accelerated inference speed. Additionally, we demonstrate the versatility of MiraMo through applications in motion transfer and video editing tasks.

cs.CV↗
arXiv2025

Self-Improvement for Audio Large Language Model using Unlabeled Speech

Recent audio LLMs have emerged rapidly, demonstrating strong generalization across various speech tasks. However, given the inherent complexity of speech signals, these models inevitably suffer from performance degradation in specific target domains. To address this, we focus on enhancing audio LLMs in target domains without any labeled data. We propose a self-improvement method called SI-SDA, leveraging the information embedded in large-model decoding to evaluate the quality of generated pseudo labels and then perform domain adaptation based on reinforcement learning optimization. Experimental results show that our method consistently and significantly improves audio LLM performance, outperforming existing baselines in WER and BLEU across multiple public datasets of automatic speech recognition (ASR), spoken question-answering (SQA), and speech-to-text translation (S2TT). Furthermore, our approach exhibits high data efficiency, underscoring its potential for real-world deployment.

cs.SD↗
arXiv2025

Analysis of multivariate event times under informative censoring using vine copula

The study of times to nonterminal events of different types and their interrelation is a compelling area of interest. The primary challenge in analyzing such multivariate event times is the presence of informative censoring by the terminal event. While numerous statistical methods have been proposed for a single nonterminal event, i.e., semi-competing risks data, there remains a dearth of tools for analyzing times to multiple nonterminal events. This article introduces a novel analysis framework that leverages the vine copula to directly estimate the joint density of multivariate times to nonterminal and terminal events. Unlike the few existing methods based on multivariate or nested copulas, the developed approach excels in capturing the heterogeneous dependence between each pair of event times (nonterminal-terminal and between-nonterminal) in terms of strength and structure. We propose a likelihood-based estimation and inference procedure, which can be implemented efficiently in sequential stages. Through extensive simulation studies, we demonstrate the satisfactory finite-sample performance of our proposed stage-wise estimators and analytical variance estimators, as well as their advantages over existing methods. We apply the developed approach to data from a crowdfunding platform to investigate the relationship between various types of creator-backer interactions and a creator's lifetime on the platform.

stat.ME↗
arXiv2025

Derivation of Tissue Properties from Basis-Vector Model Weights for Dual-Energy CT-Based Monte Carlo Proton Beam Dose Calculations

We propose a novel method, basis vector model material indexing (BVM-MI), for predicting atomic composition and mass density from two independent basis vector model weights derived from dual-energy CT (DECT) for Monte Carlo (MC) dose planning. BVM-MI employs multiple linear regression on BVM weights and their quotient to predict elemental composition and mass density for 70 representative tissues. Predicted values were imported into the TOPAS MC code to simulate proton dose deposition to a uniform cylinder phantom composed of each tissue type. The performance of BVM-MI was compared to the conventional Hounsfield Unit material indexing method (HU-MI), which estimates elemental composition and density based on CT numbers (HU). Evaluation metrics included absolute errors in predicted elemental compositions and relative percent errors in calculated mass density and mean excitation energy. Dose distributions were assessed by quantifying absolute error in the depth of 80% maximum scored dose (R80) and relative percent errors in stopping power (SP) between MC simulations using HU-MI, BVM-MI, and benchmark compositions. Lateral dose profiles were analyzed at R80 and Bragg Peak (RBP) depths for three tissues showing the largest discrepancies in R80 depth. BVM-MI outperformed HU-MI in elemental composition predictions, with mean RMSEs of 1.30% (soft tissue) and 0.1% (bony tissue), compared to 4.20% and 1.9% for HU-MI. R80 depth RMSEs were 0.2 mm (soft) and 0.1 mm (bony) for BVM-MI, vs. 1.8 mm and 0.7 mm for HU-MI. Lateral dose profile analysis showed overall smaller dose errors for BVM-MI across core, halo, and proximal aura regions. In conclusion, fully utilizing the two-parameter BVM space for material indexing led to significantly improved TOPAS MC dose calculations over the conventional HU-MI method, demonstrating BVM-MI's potential to enhance proton therapy planning.

physics.med-ph↗
arXiv2025

GenHOI: Generalizing Text-driven 4D Human-Object Interaction Synthesis for Unseen Objects

While diffusion models and large-scale motion datasets have advanced text-driven human motion synthesis, extending these advances to 4D human-object interaction (HOI) remains challenging, mainly due to the limited availability of large-scale 4D HOI datasets. In our study, we introduce GenHOI, a novel two-stage framework aimed at achieving two key objectives: 1) generalization to unseen objects and 2) the synthesis of high-fidelity 4D HOI sequences. In the initial stage of our framework, we employ an Object-AnchorNet to reconstruct sparse 3D HOI keyframes for unseen objects, learning solely from 3D HOI datasets, thereby mitigating the dependence on large-scale 4D HOI datasets. Subsequently, we introduce a Contact-Aware Diffusion Model (ContactDM) in the second stage to seamlessly interpolate sparse 3D HOI keyframes into densely temporally coherent 4D HOI sequences. To enhance the quality of generated 4D HOI sequences, we propose a novel Contact-Aware Encoder within ContactDM to extract human-object contact patterns and a novel Contact-Aware HOI Attention to effectively integrate the contact signals into diffusion models. Experimental results show that we achieve state-of-the-art results on the publicly available OMOMO and 3D-FUTURE datasets, demonstrating strong generalization abilities to unseen objects, while enabling high-fidelity 4D HOI generation.

cs.CV↗