menu

Aditya Ganeshan

CS PhD Student
Brown University
Graphics / Vision / Machine Learning / Mathematics / Music
What is my research about?

My research focuses on Symbolic Geometry, i.e. structured representations that capture how visual data is constructed, organized, or related (typically in a programmatic manner). Such representations hold the promise of making geometry interpretable, editable, and reusable across domains, yet they remain difficult to design and apply in practice. I address this challenge across three fronts: Formulation, by developing symbolic languages tailored to diverse geometric domains; Acquisition, by building learning systems that recover symbolic structure from visual data; and Manipulation, by creating learning-driven interfaces that make symbolic geometry intuitive to explore and refine. Together, these efforts aim to lower the barrier to creating new symbolic representations, enabling a Cambrian explosion of geometry languages that expand how we model, communicate, and reason about visual data.

Short Bio

I am a 5th year Ph.D. candidate in Computer Science at Brown University, advised by Daniel Ritchie. I have spent time at Adobe Research and the University of Tokyo, and previously worked at Preferred Networks in Japan and the Video Analytics Lab at IISc, India. I earned my Integrated B. Sc. and M.Sc. in Applied Mathematics from IIT Roorkee.

I am actively looking for full-time opportunities starting in 2026.
Select Research
MiGumi - Making Tightly Coupled Integral Joints Millable

A. Ganeshan, Kurt Fleischer, Wenzel Jakob, Ariel Shamir, Daniel Ritchie, Takeo Igarashi and Maria Larsson

ACM Siggraph Asia 2025, Journal at Transactions on Graphics (TOG) 2025

Pattern Analogies - Learning to Perform Programmatic Image Edits by Analogy

A. Ganeshan, Thibault Groueix, Paul Guerrero, Radomír Měch, Matthew Fisher and Daniel Ritchie

IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR) 2025

ParSEL - Parameterized Shape Editing with Language

A. Ganeshan, Ryan Y. Huang, Xianghao Xu, R. Kenny Jones and Daniel Ritchie

ACM Siggraph Asia 2024, Journal at Transactions on Graphics (TOG) 2024.

Improving Unsupervised Visual Program Inference with Code Rewriting Families

A. Ganeshan, R. Kenny Jones and Daniel Ritchie

Oral (1.8%) IEEE / CVF International Conference on Computer Vision (ICCV), 2023

Recent highlights and news:
What I am doing now

I am contributing towards a few research directions: a) building scalable text2visualprogram approach for 3D data, b) Designing tool that helps design Millable Kigumi Joints and c) Building a representation and inference system for richer procedural abstractions of articulate objects.

Most recent update: October 15th 2025.