Reconstructing articulated objects is essential for building digital twins of interactive environments. However, prior methods typically decouple geometry and motion by first reconstructing object shape in distinct states and then estimating articulation through post-hoc alignment. This separation complicates the reconstruction pipeline and restricts scalability, especially for objects with complex, multi-part articulation. We introduce a unified representation that jointly models geometry and motion using articulated 3D Gaussians. This formulation improves robustness in motion decomposition and supports articulated objects with up to 20 parts, significantly outperforming prior approaches that often struggle beyond $2-3$ parts due to brittle initialization. To systematically assess scalability and generalization, we propose MPArt-90, a new benchmark consisting of 90 articulated objects across 20 categories, each with diverse part counts and motion configurations. Extensive experiments show that our method consistently achieves superior accuracy in part-level geometry reconstruction and motion estimation across a broad range of object types. We further demonstrate applicability to downstream tasks such as robotic simulation and human-scene interaction modeling, highlighting the potential of unified articulated representations in scalable physical modeling. Project Page.
Recent advances in illumination control extend image-based methods to video, yet still facing a trade-off between lighting fidelity and temporal consistency. Moving beyond relighting, a key step toward generative modeling of real-world scenes is the joint control of camera trajectory and illumination, since visual dynamics are inherently shaped by both geometry and lighting. To this end, we present Light-X, a video generation framework that enables controllable rendering from monocular videos with both viewpoint and illumination control. 1) We propose a disentangled design that decouples geometry and lighting signals: geometry and motion are captured via dynamic point clouds projected along user-defined camera trajectories, while illumination cues are provided by a relit frame consistently projected into the same geometry. These explicit, fine-grained cues enable effective disentanglement and guide high-quality illumination. 2) To address the lack of paired multi-view and multi-illumination videos, we introduce Light-Syn, a degradation-based pipeline with inverse-mapping that synthesizes training pairs from in-the-wild monocular footage. This strategy yields a dataset covering static, dynamic, and AI-generated scenes, ensuring robust training. Extensive experiments show that Light-X outperforms baseline methods in joint camera-illumination control and surpasses prior video relighting methods under both text- and background-conditioned settings.
Universal photometric stereo (PS) is defined by two factors: it must (i) operate under arbitrary, unknown lighting conditions and (ii) avoid reliance on specific illumination models. Despite progress (e.g., SDM UniPS), two challenges remain. First, current encoders cannot guarantee that illumination and normal information are decoupled. To enforce decoupling, we introduce LINO UniPS with two key components: (i) Light Register Tokens with light alignment supervision to aggregate point, direction, and environment lights; (ii) Interleaved Attention Block featuring global cross-image attention that takes all lighting conditions together so the encoder can factor out lighting while retaining normal-related evidence. Second, high-frequency geometric details are easily lost. We address this with (i) a Wavelet-based Dual-branch Architecture and (ii) a Normal-gradient Perception Loss. These techniques yield a \textbf{unified} feature space in which lighting is explicitly represented by register tokens, while normal details are preserved via wavelet branch. We further introduce PS-Verse, a large-scale synthetic dataset graded by geometric complexity and lighting diversity, and adopt curriculum training from simple to complex scenes. Extensive experiments show new state-of-the-art results on public benchmarks (e.g., DiLiGenT, Luces), stronger generalization to real materials, and improved efficiency; ablations confirm that Light Register Tokens + Interleaved Attention Block drive better feature decoupling, while Wavelet-based Dual-branch Architecture + Normal-gradient Perception Loss recover finer details.
Abstract Introduction This study focuses on developing and evaluating a customized Generative AI chatbot designed to enhance access to large-scale educational data. The chatbot aims to assist researchers and policymakers in exploring complex datasets, such as NAEP, through natural language queries. Methods The chatbot was built using a Retrieval-Augmented Generation (RAG) framework that integrates multiple specialized agents to retrieve, interpret, and synthesize educational data. One agent was selected as a case study for performance evaluation. The study compared an automated Large Language Model (LLM)-based evaluation (“LLM-as-a-judge”) with human expert ratings to examine validity and consistency across three criteria: correctness, completeness, and communication quality. A total of 141 expert-generated questions reflecting typical user queries were used, each accompanied by a reference answer and source documentation. Chatbot’s responses were evaluated with a three-dimensional framework on Correctness, Completeness, and Communication. In addition to human evaluation, an LLM-based evaluation was implemented, and the model was provided with the rubric, human-written reference answers, and retrieved RAG contents to generate automated quality assessments. Interrater reliability among human raters and the LLM-as-a-judge were computed with quadratic weighted kappa (QWK). Results Findings showed that the LLM-as-a-judge approach achieved comparable agreement levels with human raters and demonstrated reliability across all evaluation dimensions. Interrater reliability analyses revealed no significant differences between inter-human and human-to-LLM agreement, except in the communication dimension, where human-to-LLM consistency was higher. These results indicate that the LLM-as-a-judge method can serve as a viable and consistent alternative to human evaluation for customized RAG-based chatbot assessment. Conclusions Integrating LLM-based evaluation into the assessment of Generative AI chatbots provides a scalable, reliable, and cost-effective complement to traditional human review. With human oversight for calibration and validation, this approach enables more efficient and consistent evaluation practices, advancing the use of AI tools that facilitate broader access to large-scale educational data.
Neural asset authoring and neural rendering have emerged as largely disjoint threads: one generates digital assets using neural networks for traditional graphics pipelines, while the other develops neural renderers that map conventional assets to images. However, the joint design of the asset representation and renderer remains largely unexplored. We argue that coupling them can unlock an end-to-end learnable graphics stack with benefits in fidelity, consistency, and efficiency. In this paper, we explore this possibility with NeAR : a Coupled Neural Asset–Renderer Stack. On the asset side, we build on Trellis-style Structured 3D Latents and introduce a lighting-homogenized neural asset: from a casually lit input, a rectified-flow backbone predicts a Lighting-Homogenized SLAT that encodes geometry and intrinsic material cues in a compact, view-agnostic latent. On the renderer side, we design a lighting-aware neural renderer that uses this neural asset, along with explicit view embeddings and HDR environment maps, to produce lighting-aware renderings in realtime. We validate NeAR on four tasks: (1) G-buffer–based forward rendering, (2) random-lit single-image reconstruction, (3) unknown-lit single-image relighting, and (4) novel-view relighting, where our coupled stack surpasses state-of-the-art baselines in quantitative metrics and perceptual quality. We hope this coupled asset-renderer perspective inspires new graphics stacks that view neural assets and renderers as co-designed components instead of independent ones.
