Sketches are widely used in conceptual design to externalize early ideas and communicate intent. With the rise of generative AI, sketch-to-design workflows have advanced rapidly. However, sketches are limited for organizing component-level structure and intent: parts, functions, and relations are often implicit, making systematic design space exploration difficult. We present SketchConcept, a sketch-to-design system that enables multimodal exploration through sketching and language. It allows designers to sketch out the form, then use voice or text to articulate and refine component functions and structural organization. This enables designers to explore not only satisfying appearances, but also functional and structural alternatives that are essential for design. To support this workflow, SketchConcept introduces a function-to-visual mapping mechanism that connects visual components to functional properties for component-wise iteration. We demonstrate the system through a set of representative use cases and evaluate its efficacy and usability in a two-session user study.

Generative AI (GenAI) has significantly advanced the ease and flexibility of image creation. However, it remains a challenge to precisely control spatial compositions, including object arrangement and scene conditions. To bridge this gap, we propose Canvas3D, an interactive system leveraging a 3D engine to enable precise spatial manipulation for image generation. Upon user prompt, Canvas3D automatically converts textual descriptions into interactive objects within a 3D engine-driven virtual canvas, empowering direct and precise spatial configuration. These user-defined arrangements generate explicit spatial constraints that guide generative models in accurately reflecting user intentions in the resulting images. We conducted a closed-ended comparative study between Canvas3D and a baseline system, and an open-ended, free-form study to assess overall system usability. The results indicate that Canvas3D outperforms the baseline on spatial control, interactivity, and overall user experience.

Generated Design SpaceIteration-1 Iteration-2 Iteration-3Figure 1: Exploring the design space of a desk using DesignFromX.The process begins with the user selecting a component from a reference product image-here, the legs of a wooden chair.The system identifies and suggests design features of the selected component.The user then composes this feature, in this case, the structural form, into a designated part of the desk.Based on the user-defined composition, a Generative AI model generates a new design space for the desk.In subsequent iterations, the user can incorporate additional design features from other reference products to further explore the design space while retaining the features of their previous selections.

Label differential privacy (Label DP) establishes a rigorous framework for safeguarding sensitive labels in supervised learning scenarios, operating under the premise that instance feature vectors are non-sensitive and publicly accessible. Most existing approaches predominantly rely on randomized flipping that apply uniform probability transformations to convert sensitive labels into privacy-preserved alternatives. However, these methods fail to incorporate the intrinsic structural dependency between instance-level and label-level within the privacy analysis framework, which may suffer significant utility degradation as privacy constraints more tighten. To address these limitations, we make a first attempt to explore the probabilistic generative model for Label DP, where the structured Bayesian nonparametric prior is tailored to capture dependent relationship inherent in instance space through an infinitely deep stick-breaking construction. By explicitly modeling the generative process connecting correlation representations to label distribution, we develop a tailored generative framework that enables principled privacy-utility analysis in Label DP scenarios. Specifically, by treating privacy-preserving learning as an inference task rather than a direct classification problem, our approach leverages the rich structural relationships between instances and labels, allowing for more effective privatization mechanisms that can better handle uncertainty and preserve essential statistical properties while satisfying rigorous privacy guarantees. Comprehensive experiments across multiple benchmark and real-world datasets confirm that our generative modeling framework consistently outperforms previous approaches, maintaining high utility even under strict privacy constraints where earlier methods fail. Code has been available at github11HILADP: https://github.com/hu-yc/HiLaDP.