arXiv subjects

I-Chao Shen

Publications and source records attributed to I-Chao Shen.

At least 19 recordsLinked to original sources

arXiv2026

NoiseTilt: Noise-Tilted Reverse Kernels for Diffusion Reward Alignment

We introduce the Noise-Tilted Reverse Kernel (NTRK), a reward-guided diffusion sampler that injects reward gradients through the noise term, leaving the pretrained reverse kernel unchanged and requiring only a single sample per step. Reward-guided sampling at inference time has greatly expanded the versatility of pretrained diffusion models. Yet existing methods face a trade-off. Gradient-based guidance shifts the reverse mean, steering generation but pushing intermediate states outside the region that the model was trained on and degrading quality. Search-based methods preserve quality but gain no gradient signal. No prior method achieves both. NTRK resolves this by keeping the reverse mean fixed and biasing the noise term toward high reward. This is enabled by a whitening operator, the central mechanism behind NTRK, which converts reward gradients into noise-compatible perturbations without losing their guiding signal. Across various reward alignment tasks, NTRK outperforms recent state-of-the-art baselines without losing sample quality. Remarkably, on aesthetic generation, NTRK surpasses the reward of the best baseline at 500 NFEs using only 25 NFEs, a 20 times reduction in compute.

cs.LG↗

arXiv2026

Garment Particles: A 2D--3D Symmetric Garment Representation for Generation and Editing

Practical garment design spans two modes: intuitive creation from high-level intent, such as a reference image or text description, and complex low-level editing across 2D sewing patterns and 3D draped geometry, which requires professional training to navigate their complex interdependencies. Yet existing frameworks address only part of this challenge, offering either garment generation from casual inputs or direct editing on sewing patterns. To support both ends of the spectrum, we propose Garment Particles, a 5D point-cloud representation that jointly encodes 2D sewing patterns and 3D geometry. This representation enables Garment Particles Flow (GPF), a rectified flow framework that supports intuitive generation from high-level inputs (text, images, sketches) and various editing operations on 2D sewing patterns and 3D geometries via diffusion posterior sampling. Finally, we introduce Particles-to-Pattern Flow that converts generated garment particles into curved-based patterns for simulation. We validate our model's generation ability on multiple datasets, achieving state-of-the-art garment generation results against competitive baselines. Our model also enables many garment editing scenarios, including garment interpolation, sewing pattern editing, point-cloud- and silhouette-conditioned garment generation. Our project website is at https://garment-particles.github.io .

cs.GR↗

arXiv2026

Dropping Anchor and Spherical Harmonics for Sparse-view Gaussian Splatting

Recent 3D Gaussian Splatting (3DGS) Dropout methods address overfitting under sparse-view conditions by randomly nullifying Gaussian opacities. However, we identify a neighbor compensation effect in these approaches: dropped Gaussians are often compensated by their neighbors, weakening the intended regularization. Moreover, these methods overlook the contribution of high-degree spherical harmonic coefficients (SH) to overfitting. To address these issues, we propose DropAnSH-GS, a novel anchor-based Dropout strategy. Rather than dropping Gaussians independently, our method randomly selects certain Gaussians as anchors and simultaneously removes their spatial neighbors. This effectively disrupts local redundancies near anchors and encourages the model to learn more robust, globally informed representations. Furthermore, we extend the Dropout to color attributes by randomly dropping higher-degree SH to concentrate appearance information in lower-degree SH. This strategy further mitigates overfitting and enables flexible post-training model compression via SH truncation. Experimental results demonstrate that DropAnSH-GS substantially outperforms existing Dropout methods with negligible computational overhead, and can be readily integrated into various 3DGS variants to enhance their performances. Project Website: https://sk-fun.fun/DropAnSH-GS

cs.CV↗

arXiv2025

Cascaded Robust Rectification for Arbitrary Document Images

Document rectification in real-world scenarios poses significant challenges due to extreme variations in camera perspectives and physical distortions. Driven by the insight that complex transformations can be decomposed and resolved progressively, we introduce a novel multi-stage framework that progressively reverses distinct distortion types in a coarse-to-fine manner. Specifically, our framework first performs a global affine transformation to correct perspective distortions arising from the camera's viewpoint, then rectifies geometric deformations resulting from physical paper curling and folding, and finally employs a content-aware iterative process to eliminate fine-grained content distortions. To address limitations in existing evaluation protocols, we also propose two enhanced metrics: layout-aligned OCR metrics (AED/ACER) for a stable assessment that decouples geometric rectification quality from the layout analysis errors of OCR engines, and masked AD/AAD (AD-M/AAD-M) tailored for accurately evaluating geometric distortions in documents with incomplete boundaries. Extensive experiments show that our method establishes new state-of-the-art performance on multiple challenging benchmarks, yielding a substantial reduction of 14.1\%--34.7\% in the AAD metric and demonstrating superior efficacy in real-world applications. The code will be publicly available at https://github.com/chaoyunwang/ArbDR.

cs.CV↗

arXiv2025

Axis-Aligned Document Dewarping

Document dewarping is crucial for many applications. However, existing learning-based methods rely heavily on supervised regression with annotated data without fully leveraging the inherent geometric properties of physical documents. Our key insight is that a well-dewarped document is defined by its axis-aligned feature lines. This property aligns with the inherent axis-aligned nature of the discrete grid geometry in planar documents. Harnessing this property, we introduce three synergistic contributions: for the training phase, we propose an axis-aligned geometric constraint to enhance document dewarping; for the inference phase, we propose an axis alignment preprocessing strategy to reduce the dewarping difficulty; and for the evaluation phase, we introduce a new metric, Axis-Aligned Distortion (AAD), that not only incorporates geometric meaning and aligns with human visual perception but also demonstrates greater robustness. As a result, our method achieves state-of-the-art performance on multiple existing benchmarks, improving the AAD metric by 18.2% to 34.5%. The code is publicly available at https://github.com/chaoyunwang/AADD.

cs.CV↗

arXiv2025

Real-Time Per-Garment Virtual Try-On with Temporal Consistency for Loose-Fitting Garments

Per-garment virtual try-on methods collect garment-specific datasets and train networks tailored to each garment to achieve superior results. However, these approaches often struggle with loose-fitting garments due to two key limitations: (1) They rely on human body semantic maps to align garments with the body, but these maps become unreliable when body contours are obscured by loose-fitting garments, resulting in degraded outcomes; (2) They train garment synthesis networks on a per-frame basis without utilizing temporal information, leading to noticeable jittering artifacts. To address the first limitation, we propose a two-stage approach for robust semantic map estimation. First, we extract a garment-invariant representation from the raw input image. This representation is then passed through an auxiliary network to estimate the semantic map. This enhances the robustness of semantic map estimation under loose-fitting garments during garment-specific dataset generation. To address the second limitation, we introduce a recurrent garment synthesis framework that incorporates temporal dependencies to improve frame-to-frame coherence while maintaining real-time performance. We conducted qualitative and quantitative evaluations to demonstrate that our method outperforms existing approaches in both image quality and temporal coherence. Ablation studies further validate the effectiveness of the garment-invariant representation and the recurrent synthesis framework.

cs.GR↗

arXiv2025

LayoutRectifier: An Optimization-based Post-processing for Graphic Design Layout Generation

Recent deep learning methods can generate diverse graphic design layouts efficiently. However, these methods often create layouts with flaws, such as misalignment, unwanted overlaps, and unsatisfied containment. To tackle this issue, we propose an optimization-based method called LayoutRectifier, which gracefully rectifies auto-generated graphic design layouts to reduce these flaws while minimizing deviation from the generated layout. The core of our method is a two-stage optimization. First, we utilize grid systems, which professional designers commonly use to organize elements, to mitigate misalignments through discrete search. Second, we introduce a novel box containment function designed to adjust the positions and sizes of the layout elements, preventing unwanted overlapping and promoting desired containment. We evaluate our method on content-agnostic and content-aware layout generation tasks and achieve better-quality layouts that are more suitable for downstream graphic design tasks. Our method complements learning-based layout generation methods and does not require additional training.

cs.GR↗

arXiv2025

AutoSketch: VLM-assisted Style-Aware Vector Sketch Completion

The ability to automatically complete a partial sketch that depicts a complex scene, e.g., "a woman chatting with a man in the park", is very useful. However, existing sketch generation methods create sketches from scratch; they do not complete a partial sketch in the style of the original. To address this challenge, we introduce AutoSketch, a styleaware vector sketch completion method that accommodates diverse sketch styles. Our key observation is that the style descriptions of a sketch in natural language preserve the style during automatic sketch completion. Thus, we use a pretrained vision-language model (VLM) to describe the styles of the partial sketches in natural language and replicate these styles using newly generated strokes. We initially optimize the strokes to match an input prompt augmented by style descriptions extracted from the VLM. Such descriptions allow the method to establish a diffusion prior in close alignment with that of the partial sketch. Next, we utilize the VLM to generate an executable style adjustment code that adjusts the strokes to conform to the desired style. We compare our method with existing methods across various sketch styles and prompts, performed extensive ablation studies and qualitative and quantitative evaluations, and demonstrate that AutoSketch can support various sketch scenarios.

cs.CV↗

arXiv2025

MeshLLM: Empowering Large Language Models to Progressively Understand and Generate 3D Mesh

We present MeshLLM, a novel framework that leverages large language models (LLMs) to understand and generate text-serialized 3D meshes. Our approach addresses key limitations in existing methods, including the limited dataset scale when catering to LLMs' token length and the loss of 3D structural information during mesh serialization. We introduce a Primitive-Mesh decomposition strategy, which divides 3D meshes into structurally meaningful subunits. This enables the creation of a large-scale dataset with 1500k+ samples, almost 50 times larger than previous methods, which aligns better with the LLM scaling law principles. Furthermore, we propose inferring face connectivity from vertices and local mesh assembly training strategies, significantly enhancing the LLMs' ability to capture mesh topology and spatial structures. Experiments show that MeshLLM outperforms the state-of-the-art LLaMA-Mesh in both mesh generation quality and shape understanding, highlighting its great potential in processing text-serialized 3D meshes.

cs.GR↗

arXiv2025

NeRF Is a Valuable Assistant for 3D Gaussian Splatting

We introduce NeRF-GS, a novel framework that jointly optimizes Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS). This framework leverages the inherent continuous spatial representation of NeRF to mitigate several limitations of 3DGS, including sensitivity to Gaussian initialization, limited spatial awareness, and weak inter-Gaussian correlations, thereby enhancing its performance. In NeRF-GS, we revisit the design of 3DGS and progressively align its spatial features with NeRF, enabling both representations to be optimized within the same scene through shared 3D spatial information. We further address the formal distinctions between the two approaches by optimizing residual vectors for both implicit features and Gaussian positions to enhance the personalized capabilities of 3DGS. Experimental results on benchmark datasets show that NeRF-GS surpasses existing methods and achieves state-of-the-art performance. This outcome confirms that NeRF and 3DGS are complementary rather than competing, offering new insights into hybrid approaches that combine 3DGS and NeRF for efficient 3D scene representation.

cs.CV↗

arXiv2025

Low-Barrier Dataset Collection with Real Human Body for Interactive Per-Garment Virtual Try-On

Existing image-based virtual try-on methods are often limited to the front view and lack real-time performance. While per-garment virtual try-on methods have tackled these issues by capturing per-garment datasets and training per-garment neural networks, they still encounter practical limitations: (1) the robotic mannequin used to capture per-garment datasets is prohibitively expensive for widespread adoption and fails to accurately replicate natural human body deformation; (2) the synthesized garments often misalign with the human body. To address these challenges, we propose a low-barrier approach for collecting per-garment datasets using real human bodies, eliminating the necessity for a customized robotic mannequin. We also introduce a hybrid person representation that enhances the existing intermediate representation with a simplified DensePose map. This ensures accurate alignment of synthesized garment images with the human body and enables human-garment interaction without the need for customized wearable devices. We performed qualitative and quantitative evaluations against other state-of-the-art image-based virtual try-on methods and conducted ablation studies to demonstrate the superiority of our method regarding image quality and temporal consistency. Finally, our user study results indicated that most participants found our virtual try-on system helpful for making garment purchasing decisions.

cs.GR↗

arXiv2025

GarmentImage: Raster Encoding of Garment Sewing Patterns with Diverse Topologies

Garment sewing patterns are the design language behind clothing, yet their current vector-based digital representations weren't built with machine learning in mind. Vector-based representation encodes a sewing pattern as a discrete set of panels, each defined as a sequence of lines and curves, stitching information between panels and the placement of each panel around a body. However, this representation causes two major challenges for neural networks: discontinuity in latent space between patterns with different topologies and limited generalization to garments with unseen topologies in the training data. In this work, we introduce GarmentImage, a unified raster-based sewing pattern representation. GarmentImage encodes a garment sewing pattern's geometry, topology and placement into multi-channel regular grids. Machine learning models trained on GarmentImage achieve seamless transitions between patterns with different topologies and show better generalization capabilities compared to models trained on vector-based representation. We demonstrate the effectiveness of GarmentImage across three applications: pattern exploration in latent space, text-based pattern editing, and image-to-pattern prediction. The results show that GarmentImage achieves superior performance on these applications using only simple convolutional networks.

cs.GR↗

arXiv2025

FontCraft: Multimodal Font Design Using Interactive Bayesian Optimization

Creating new fonts requires a lot of human effort and professional typographic knowledge. Despite the rapid advancements of automatic font generation models, existing methods require users to prepare pre-designed characters with target styles using font-editing software, which poses a problem for non-expert users. To address this limitation, we propose FontCraft, a system that enables font generation without relying on pre-designed characters. Our approach integrates the exploration of a font-style latent space with human-in-the-loop preferential Bayesian optimization and multimodal references, facilitating efficient exploration and enhancing user control. Moreover, FontCraft allows users to revisit previous designs, retracting their earlier choices in the preferential Bayesian optimization process. Once users finish editing the style of a selected character, they can propagate it to the remaining characters and further refine them as needed. The system then generates a complete outline font in OpenType format. We evaluated the effectiveness of FontCraft through a user study comparing it to a baseline interface. Results from both quantitative and qualitative evaluations demonstrate that FontCraft enables non-expert users to design fonts efficiently.

cs.HC↗

arXiv2024

AvatarPerfect: User-Assisted 3D Gaussian Splatting Avatar Refinement with Automatic Pose Suggestion

Creating high-quality 3D avatars using 3D Gaussian Splatting (3DGS) from a monocular video benefits virtual reality and telecommunication applications. However, existing automatic methods exhibit artifacts under novel poses due to limited information in the input video. We propose AvatarPerfect, a novel system that allows users to iteratively refine 3DGS avatars by manually editing the rendered avatar images. In each iteration, our system suggests a new body and camera pose to help users identify and correct artifacts. The edited images are then used to update the current avatar, and our system suggests the next body and camera pose for further refinement. To investigate the effectiveness of AvatarPerfect, we conducted a user study comparing our method to an existing 3DGS editor SuperSplat, which allows direct manipulation of Gaussians without automatic pose suggestions. The results indicate that our system enables users to obtain higher quality refined 3DGS avatars than the existing 3DGS editor.

cs.HC↗

arXiv2024

FontCLIP: A Semantic Typography Visual-Language Model for Multilingual Font Applications

Acquiring the desired font for various design tasks can be challenging and requires professional typographic knowledge. While previous font retrieval or generation works have alleviated some of these difficulties, they often lack support for multiple languages and semantic attributes beyond the training data domains. To solve this problem, we present FontCLIP: a model that connects the semantic understanding of a large vision-language model with typographical knowledge. We integrate typography-specific knowledge into the comprehensive vision-language knowledge of a pretrained CLIP model through a novel finetuning approach. We propose to use a compound descriptive prompt that encapsulates adaptively sampled attributes from a font attribute dataset focusing on Roman alphabet characters. FontCLIP's semantic typographic latent space demonstrates two unprecedented generalization abilities. First, FontCLIP generalizes to different languages including Chinese, Japanese, and Korean (CJK), capturing the typographical features of fonts across different languages, even though it was only finetuned using fonts of Roman characters. Second, FontCLIP can recognize the semantic attributes that are not presented in the training data. FontCLIP's dual-modality and generalization abilities enable multilingual and cross-lingual font retrieval and letter shape optimization, reducing the burden of obtaining desired fonts.

cs.CV↗

arXiv2023

EvIcon: Designing High-Usability Icon with Human-in-the-loop Exploration and IconCLIP

Interface icons are prevalent in various digital applications. Due to limited time and budgets, many designers rely on informal evaluation, which often results in poor usability icons. In this paper, we propose a unique human-in-the-loop framework that allows our target users, i.e., novice and professional UI designers, to improve the usability of interface icons efficiently. We formulate several usability criteria into a perceptual usability function and enable users to iteratively revise an icon set with an interactive design tool, EvIcon. We take a large-scale pre-trained joint image-text embedding (CLIP) and fine-tune it to embed icon visuals with icon tags in the same embedding space (IconCLIP). During the revision process, our design tool provides two types of instant perceptual usability feedback. First, we provide perceptual usability feedback modeled by deep learning models trained on IconCLIP embeddings and crowdsourced perceptual ratings. Second, we use the embedding space of IconCLIP to assist users in improving icons' visual distinguishability among icons within the user-prepared icon set. To provide the perceptual prediction, we compiled IconCEPT10K, the first large-scale dataset of perceptual usability ratings over $10,000$ interface icons, by conducting a crowdsourcing study. We demonstrated that our framework could benefit UI designers' interface icon revision process with a wide range of professional experience. Moreover, the interface icons designed using our framework achieved better semantic distance and familiarity, verified by an additional online user study.

cs.GR↗

arXiv2022

NeRF-In: Free-Form NeRF Inpainting with RGB-D Priors

Though Neural Radiance Field (NeRF) demonstrates compelling novel view synthesis results, it is still unintuitive to edit a pre-trained NeRF because the neural network's parameters and the scene geometry/appearance are often not explicitly associated. In this paper, we introduce the first framework that enables users to remove unwanted objects or retouch undesired regions in a 3D scene represented by a pre-trained NeRF without any category-specific data and training. The user first draws a free-form mask to specify a region containing unwanted objects over a rendered view from the pre-trained NeRF. Our framework first transfers the user-provided mask to other rendered views and estimates guiding color and depth images within these transferred masked regions. Next, we formulate an optimization problem that jointly inpaints the image content in all masked regions across multiple views by updating the NeRF model's parameters. We demonstrate our framework on diverse scenes and show it obtained visual plausible and structurally consistent results across multiple views using shorter time and less user manual efforts.

cs.CV↗

arXiv2022

StylePart: Image-based Shape Part Manipulation

Due to a lack of image-based "part controllers", shape manipulation of man-made shape images, such as resizing the backrest of a chair or replacing a cup handle is not intuitive. To tackle this problem, we present StylePart, a framework that enables direct shape manipulation of an image by leveraging generative models of both images and 3D shapes. Our key contribution is a shape-consistent latent mapping function that connects the image generative latent space and the 3D man-made shape attribute latent space. Our method "forwardly maps" the image content to its corresponding 3D shape attributes, where the shape part can be easily manipulated. The attribute codes of the manipulated 3D shape are then "backwardly mapped" to the image latent code to obtain the final manipulated image. We demonstrate our approach through various manipulation tasks, including part replacement, part resizing, and viewpoint manipulation, and evaluate its effectiveness through extensive ablation studies.

cs.CV↗