Lingyun Yu

Yuhao Sun, Lingyun Yu, Haoxiang Xu et al.

Concept erasure has emerged as a promising approach to mitigate undesired or unsafe content in diffusion models, yet existing methods still face significant limitations. While training-based methods are effective, their high computational cost limits scalability. Editing-based methods are more efficient and deployment-friendly, yet they struggle to simultaneously achieve precise concept erasure and preserve overall generative capacity. We identify this core limitation of the editing-based methods as reliance on additive parameter updates. Our empirical analysis reveals that concept semantics primarily depend on neuron direction rather than neuron magnitude, while overall generative capacity relies on the angular geometry of neurons. As additive updates inherently entangle direction, magnitude, and angular geometry, they inevitably introduce unintended interference between concept erasure and overall generation performance. To address this, we propose Orthogonal Concept Erasure (OCE), which reformulates editing-based erasure as multiplicative parameter updates from a geometric perspective. Specifically, OCE applies layer-wise orthogonal transformations derived from a closed-form solution to the parameters, enabling precise concept erasure while preserving the neuron magnitude and angular geometry. Furthermore, to address conflicting constraints in multi-concept erasure, OCE introduces a subspace-level objective with structured subspace manipulation, yielding a more effective and scalable erasure. Extensive experiments on single- and multi-concept erasure demonstrate that OCE outperforms existing methods in concept erasure and non-target preservation, erasing up to 100 concepts in 4.3 s. Code: https://github.com/HansSunY/OCE.

Quanwei Yang, Xinchen Liu, Wu Liu et al.

Human Video Motion Transfer (HVMT) aims to, given an image of a source person, generate his/her video that imitates the motion of the driving person. Existing methods for HVMT mainly exploit Generative Adversarial Networks (GANs) to perform the warping operation based on the flow estimated from the source person image and each driving video frame. However, these methods always generate obvious artifacts due to the dramatic differences in poses, scales, and shifts between the source person and the driving person. To overcome these challenges, this paper presents a novel REgionto-whole human MOtion Transfer (REMOT) framework based on GANs. To generate realistic motions, the REMOT adopts a progressive generation paradigm: it first generates each body part in the driving pose without flow-based warping, then composites all parts into a complete person of the driving motion. Moreover, to preserve the natural global appearance, we design a Global Alignment Module to align the scale and position of the source person with those of the driving person based on their layouts. Furthermore, we propose a Texture Alignment Module to keep each part of the person aligned according to the similarity of the texture. Finally, through extensive quantitative and qualitative experiments, our REMOT achieves state-of-the-art results on two public benchmarks.

Jean-Daniel Fekete, Yifan Hu, Dominik Moritz et al.

The rapid advancement of AI is transforming human-centered systems, with profound implications for human-AI interaction, human-data interaction, and visual analytics. In the AI era, data analysis increasingly involves large-scale, heterogeneous, and multimodal data that is predominantly unstructured, as well as foundation models such as LLMs and VLMs, which introduce additional uncertainty into analytical processes. These shifts expose persistent challenges for human-data interactive systems, including perceptually misaligned latency, scalability constraints, limitations of existing interaction and exploration paradigms, and growing uncertainty regarding the reliability and interpretability of AI-generated insights. Responding to these challenges requires moving beyond conventional efficiency and scalability metrics, redefining the roles of humans and machines in analytical workflows, and incorporating cognitive, perceptual, and design principles into every level of the human-data interaction stack. This paper investigates the challenges introduced by recent advances in AI and examines how these developments are reshaping the ways users engage with data, while outlining limitations and open research directions for building human-centered AI systems for interactive data analysis in the AI era.

Yuhao Sun, Lingyun Yu, Hongtao Xie et al.

With the rapid development of face recognition (FR) systems, the privacy of face images on social media is facing severe challenges due to the abuse of unauthorized FR systems. Some studies utilize adversarial attack techniques to defend against malicious FR systems by generating adversarial examples. However, the generated adversarial examples, i.e., the protected face images, tend to suffer from subpar visual quality and low transferability. In this paper, we propose a novel face protection approach, dubbed DiffAM, which leverages the powerful generative ability of diffusion models to generate high-quality protected face images with adversarial makeup transferred from reference images. To be specific, we first introduce a makeup removal module to generate non-makeup images utilizing a fine-tuned diffusion model with guidance of textual prompts in CLIP space. As the inverse process of makeup transfer, makeup removal can make it easier to establish the deterministic relationship between makeup domain and non-makeup domain regardless of elaborate text prompts. Then, with this relationship, a CLIP-based makeup loss along with an ensemble attack strategy is introduced to jointly guide the direction of adversarial makeup domain, achieving the generation of protected face images with natural-looking makeup and high black-box transferability. Extensive experiments demonstrate that DiffAM achieves higher visual quality and attack success rates with a gain of 12.98% under black-box setting compared with the state of the arts. The code will be available at https://github.com/HansSunY/DiffAM.

Peng Jia, Zhen Xiao, Jia Li et al.

High-quality, real-time talking head synthesis remains a fundamental challenge in computer vision. Existing reconstruction- and rendering-based methods typically rely on identity-specific models, limiting cross-identity generalization. To address this issue, we propose SDTalk, a one-shot 3D Gaussian Splatting (3DGS)-based framework that generalizes to unseen identities without personalized training or fine-tuning. Our framework comprises two modules with a two-stage training strategy. In the first stage, we incorporate structured facial priors into the reconstruction module and separately predict 3DGS parameters for visible and occluded regions, enabling complete head reconstruction from a single image. In the second stage, we introduce a dual-branch motion field to model coarse and fine facial dynamics, improving detail fidelity and lip synchronization. Experiments demonstrate that SDTalk surpasses existing methods in both visual quality and inference efficiency.

Yuhao Sun, Zhuoer Xu, Shiwen Cui et al.

Large Language Models (LLMs) have achieved remarkable progress across a wide range of tasks, but remain vulnerable to safety risks such as harmful content generation and jailbreak attacks. Existing safety techniques -- including external guardrails, inference-time guidance, and post-training alignment -- each face limitations in balancing safety, utility, and controllability. In this work, we propose UpSafe$^\circ$C, a unified framework for enhancing LLM safety through safety-aware upcycling. Our approach first identifies safety-critical layers and upcycles them into a sparse Mixture-of-Experts (MoE) structure, where the router acts as a soft guardrail that selectively activates original MLPs and added safety experts. We further introduce a two-stage SFT strategy to strengthen safety discrimination while preserving general capabilities. To enable flexible control at inference time, we introduce a safety temperature mechanism, allowing dynamic adjustment of the trade-off between safety and utility. Experiments across multiple benchmarks, base model, and model scales demonstrate that UpSafe$^\circ$C achieves robust safety improvements against harmful and jailbreak inputs, while maintaining competitive performance on general tasks. Moreover, analysis shows that safety temperature provides fine-grained inference-time control that achieves the Pareto-optimal frontier between utility and safety. Our results highlight a new direction for LLM safety: moving from static alignment toward dynamic, modular, and inference-aware control.

Wanfang Xu, Lixiang Zhao, Haiwen Song et al.

In this work, we introduce Mozualization, a music generation and editing tool that creates multi-style embedded music by integrating diverse inputs, such as keywords, images, and sound clips (e.g., segments from various pieces of music or even a playful cat's meow). Our work is inspired by the ways people express their emotions -- writing mood-descriptive poems or articles, creating drawings with warm or cool tones, or listening to sad or uplifting music. Building on this concept, we developed a tool that transforms these emotional expressions into a cohesive and expressive song, allowing users to seamlessly incorporate their unique preferences and inspirations. To evaluate the tool and, more importantly, gather insights for its improvement, we conducted a user study involving nine music enthusiasts. The study assessed user experience, engagement, and the impact of interacting with and listening to the generated music.

Lei Chen, Yilin Liu, Yue Li et al.

Visual cues are essential in computer-mediated communication. It is especially important when communication happens in a collaboration scenario that requires focusing several users' attention on aspecific object among other similar ones. This paper explores the effect of visual cues on pointing tasks in co-located Augmented Reality (AR) collaboration. A user study (N = 32, 16 pairs) was conducted to compare two types of visual cues: Pointing Line (PL)and Moving Track (MT). Both are head-based visual techniques.Through a series of collaborative pointing tasks on objects with different states (static and dynamic) and density levels (low, mediumand high), the results showed that PL was better on task performance and usability, but MT was rated higher on social presenceand user preference. Based on our results, some design implicationsare provided for pointing tasks in co-located AR collaboration.

Tan Tang, Yanhong Wu, Lingyun Yu et al.

Video moderation, which refers to remove deviant or explicit content from e-commerce livestreams, has become prevalent owing to social and engaging features. However, this task is tedious and time consuming due to the difficulties associated with watching and reviewing multimodal video content, including video frames and audio clips. To ensure effective video moderation, we propose VideoModerator, a risk-aware framework that seamlessly integrates human knowledge with machine insights. This framework incorporates a set of advanced machine learning models to extract the risk-aware features from multimodal video content and discover potentially deviant videos. Moreover, this framework introduces an interactive visualization interface with three views, namely, a video view, a frame view, and an audio view. In the video view, we adopt a segmented timeline and highlight high-risk periods that may contain deviant information. In the frame view, we present a novel visual summarization method that combines risk-aware features and video context to enable quick video navigation. In the audio view, we employ a storyline-based design to provide a multi-faceted overview which can be used to explore audio content. Furthermore, we report the usage of VideoModerator through a case scenario and conduct experiments and a controlled user study to validate its effectiveness.

Junxiu Tang, Lingyun Yu, Tan Tang et al.

Transitions are widely used in data videos to seamlessly connect data-driven charts or connect visualizations and non-data-driven motion graphics. To inform the transition designs in data videos, we conduct a content analysis based on more than 3500 clips extracted from 284 data videos. We annotate visualization types and transition designs on these segments, and examine how these transitions help make connections between contexts. We propose a taxonomy of transitions in data videos, where two transition categories are defined in building fluent narratives by using visual variables.

Tan Tang, Renzhong Li, Xinke Wu et al.

Storyline visualizations are an effective means to present the evolution of plots and reveal the scenic interactions among characters. However, the design of storyline visualizations is a difficult task as users need to balance between aesthetic goals and narrative constraints. Despite that the optimization-based methods have been improved significantly in terms of producing aesthetic and legible layouts, the existing (semi-) automatic methods are still limited regarding 1) efficient exploration of the storyline design space and 2) flexible customization of storyline layouts. In this work, we propose a reinforcement learning framework to train an AI agent that assists users in exploring the design space efficiently and generating well-optimized storylines. Based on the framework, we introduce PlotThread, an authoring tool that integrates a set of flexible interactions to support easy customization of storyline visualizations. To seamlessly integrate the AI agent into the authoring process, we employ a mixed-initiative approach where both the agent and designers work on the same canvas to boost the collaborative design of storylines. We evaluate the reinforcement learning model through qualitative and quantitative experiments and demonstrate the usage of PlotThread using a collection of use cases.

Jun Yuan, Shouxing Xiang, Jiazhi Xia et al.

Given a scatterplot with tens of thousands of points or even more, a natural question is which sampling method should be used to create a small but "good" scatterplot for a better abstraction. We present the results of a user study that investigates the influence of different sampling strategies on multi-class scatterplots. The main goal of this study is to understand the capability of sampling methods in preserving the density, outliers, and overall shape of a scatterplot. To this end, we comprehensively review the literature and select seven typical sampling strategies as well as eight representative datasets. We then design four experiments to understand the performance of different strategies in maintaining: 1) region density; 2) class density; 3) outliers; and 4) overall shape in the sampling results. The results show that: 1) random sampling is preferred for preserving region density; 2) blue noise sampling and random sampling have comparable performance with the three multi-class sampling strategies in preserving class density; 3) outlier biased density based sampling, recursive subdivision based sampling, and blue noise sampling perform the best in keeping outliers; and 4) blue noise sampling outperforms the others in maintaining the overall shape of a scatterplot.

Chen Zhu-Tian, Wei Zeng, Zhiguang Yang et al.

Selection is a fundamental task in exploratory analysis and visualization of 3D point clouds. Prior researches on selection methods were developed mainly based on heuristics such as local point density, thus limiting their applicability in general data. Specific challenges root in the great variabilities implied by point clouds (e.g., dense vs. sparse), viewpoint (e.g., occluded vs. non-occluded), and lasso (e.g., small vs. large). In this work, we introduce LassoNet, a new deep neural network for lasso selection of 3D point clouds, attempting to learn a latent mapping from viewpoint and lasso to point cloud regions. To achieve this, we couple user-target points with viewpoint and lasso information through 3D coordinate transform and naive selection, and improve the method scalability via an intention filtering and farthest point sampling. A hierarchical network is trained using a dataset with over 30K lasso-selection records on two different point cloud data. We conduct a formal user study to compare LassoNet with two state-of-the-art lasso-selection methods. The evaluations confirm that our approach improves the selection effectiveness and efficiency across different combinations of 3D point clouds, viewpoints, and lasso selections. Project Website: https://lassonet.github.io