Publications
For a full overview of my publications, visit my Google Scholar page.
3D Generative AI with Physical Constraints
I develop generative AI systems that integrate physical constraints—such as structural integrity, mechanical behavior, and material properties—into 3D model generation. My work bridges the gap between visually compelling AI-generated models and physically functional objects that can be successfully manufactured and used in the real world.
MechStyle: Augmenting Generative AI with Mechanical Simulation to Create Stylized and Structurally Viable 3D Models
ACM SCF 2025
We present MechStyle, a system that integrates mechanical simulation into generative 3D workflows to create objects that are both visually compelling and physically functional. By combining generative AI with finite element analysis, users can generate stylized 3D models that maintain structural integrity and can be successfully 3D printed.
TactStyle: Generating Tactile Textures with Generative AI for Digital Fabrication
CHI 2025
TactStyle enables users to generate tactile textures from visual images using generative AI, transforming 2D images into 3D printable tactile reliefs. The system encodes texture and depth information, making visual content accessible through touch for blind and low-vision users.
Style2Fab: Functionality-Aware Segmentation for Fabricating Personalized 3D Models with Generative AI
UIST 2023
Style2Fab enables users to personalize 3D models generated by AI while preserving their functional properties. The system uses functionality-aware segmentation to separate stylistic and functional regions, allowing users to modify aesthetics without compromising structural integrity.
Text2Texture: Generating 3D-Printed Models with Textures based on Text and Image Prompts
UIST 2025 (Adjunct Proceedings)
Text2Texture is a webtool that converts 2D color images / paintings into textured 3D models, allowing users to create tactile 3D models from their favorite artworks. The system extracts depth information using a monocular estimator and local texture information using a stable diffusion model, superimposing macro- and micro-scale geometries to produce composite 3D models with color, depth, and texture.
Accessibility & Inclusive Design
I create accessible design tools and systems that enable users with visual impairments to participate in 3D design and fabrication. My work focuses on making digital fabrication more inclusive through AI-assisted interfaces and tactile representations of visual content.
A11yShape: AI-Assisted 3-D Modeling for Blind and Low-Vision Programmers
ASSETS 2025
A11yShape is an AI-assisted 3D modeling system designed for blind and low-vision programmers. The system enables users to create and manipulate 3D models through accessible interfaces, making 3D design and fabrication more inclusive for users with visual impairments.
Mixed Reality & 3D Design with AI
I explore how to capture and reconstruct the interactivity of physical objects in mixed reality environments. My work enables users to preserve and interact with meaningful personal items digitally, creating rich mixed reality experiences that maintain both the appearance and functional behaviors of physical artifacts.
InteRecon: Towards Reconstructing Interactivity of Personal Memorable Items in Mixed Reality
CHI 2025
InteRecon reconstructs the interactivity of personal memorable items in mixed reality, allowing users to preserve and interact with meaningful objects digitally. The system captures both the physical appearance and functional behaviors of objects, enabling rich mixed reality experiences with personal artifacts.
Computational Design & Fabrication Tools
I develop computational tools and systems that enhance the 3D printing and digital fabrication workflow. My work includes design toolkits for specialized fabrication processes, repository augmentation for better model discovery, novel printing technologies, and authentication systems for digital manufacturing.
WireBend-kit: A Computational Design and Fabrication Toolkit for Wirebending Custom 3D Wireframe Structures
ACM SCF 2025
WireBend-kit provides a computational design and fabrication toolkit for creating custom 3D wireframe structures through wirebending. The system enables users to design complex wireframe geometries and automatically generates fabrication instructions for bending wire into desired shapes.
SliceHub: Augmenting Shared 3D Model Repositories with Slicing Results for 3D Printing
CHI 2022 (Adjunct Proceedings)
SliceHub augments shared 3D model repositories with slicing results, providing users with previews of how models will print before downloading. The system helps users identify printability issues and select appropriate models by showing slicing outcomes directly in the repository interface.
ChromoPrint: A Multi-Color 3D Printer Based on a Reprogrammable Photochromic Resin
CHI 2022 (Adjunct Proceedings)
ChromoPrint is a multi-color 3D printer that uses reprogrammable photochromic resin to create objects with dynamic color properties. The system enables post-printing color changes through selective UV exposure, allowing users to customize and recolor 3D printed objects after fabrication.
G-ID: Identifying 3D Prints Using Slicing Parameters
CHI 2020
G-ID identifies 3D printed objects by analyzing their slicing parameters, creating a unique fingerprint based on the manufacturing process. The system enables authentication and tracking of 3D printed objects, addressing security and intellectual property concerns in digital fabrication.
Programmable Materials & Interfaces
I work with programmable materials, particularly magnetic pixels, to create reconfigurable physical interfaces and adaptive assembly systems. My research enables rapid prototyping of interactive interfaces and robotic systems that can dynamically change their behavior through material programming.
Selective Self-Assembly using Re-Programmable Magnetic Pixels
IROS 2023
This work presents a system for selective self-assembly using re-programmable magnetic pixels. Objects can be dynamically reconfigured through magnetic programming, enabling flexible and adaptive robotic assembly processes that can change behavior based on task requirements.
Mixels: Fabricating Interfaces using Programmable Magnetic Pixels
UIST 2022
Mixels introduces programmable magnetic pixels that can be used to fabricate interactive interfaces. The system enables rapid prototyping of physical interfaces with reconfigurable magnetic properties, allowing for dynamic and adaptable user interfaces in physical computing applications.