About

Venkataram Sivaram

I am a first-year PhD student at the Computer Science and Artificial Intelligence Lab (CSAIL) at Massachusetts Institute of Technology (MIT), advised by Professors Fredo Durand and Jonathan-Ragan Kelley.

I graduated from UC San Diego (UCSD) in Spring of 2025 with a Bachelor of Science in Computer Science, earning the Honors with highest distinction and a minor in Mathematics. During this time, I was actively involved in Computer Graphics research at the Center for Visual Computing, where I was fortunate to be mentored by Professors Ravi Ramamoorthi and Tzu-Mao Li.

My research interests span various topics in computer graphics, including neural graphics, differentiable rendering, and appearance modeling. I'm passionate about developing efficient algorithms for 3D graphics and am eagerly exploring the intersection of machine learning and computer graphics.

Previously, I was an intern at NVIDIA during Summer 2024 and 2025, where I worked with the Slang development team. I was also a teaching assistance for CSE 167 at UCSD in Fall 2024 and Winter 2025.

I have reviewed multiple articles for ACM SIGGRAPH and ACM TOG.

Education

PhD, Computer Science

Fall 2025 to Present
Massachusetts Institute of Technology, CSAIL

Bachelor of Science, Computer Science

Fall 2022 to Spring 2025
UC San Diego, CSE
Honors with highest distinction, Minor in Mathematics

Publications

Modeling and Rendering Glow Discharge
SIGGRAPH 2025 (Conference)

Modeling and Rendering Glow Discharge

Venkataram Sivaram / Ravi Ramamoorthi / Tzu-Mao Li
Introduces a fast, approximate model for glow discharge, enabling interactive Monte Carlo rendering of emissive elements such as Neon lights.
Geometry Field Splatting with Gaussian Surfels
CVPR 2025

Geometry Field Splatting with Gaussian Surfels

Kaiwen Jiang / Venkataram Sivaram / Cheng Peng / Ravi Ramamoorthi
Accurate geometric reconstruction of opaque surfaces from images using Gaussian surfels and a nearly exact differentiable renderer.
Neural Geometry Fields for Meshes
SIGGRAPH 2024 (Conference)

Neural Geometry Fields for Meshes

Venkataram Sivaram / Tzu-Mao Li / Ravi Ramamoorthi
A neural representation for 3D mesh surfaces using coarse quadrangular patches and coordinate neural networks for detail refinement.
Parameter-space ReSTIR for Differentiable and Inverse Rendering
SIGGRAPH 2023 (Conference)

Parameter-space ReSTIR for Differentiable and Inverse Rendering

Wesley Chang / Venkataram Sivaram / Derek Nowrouzezahrai / Toshiya Hachisuka / Ravi Ramamoorthi / Tzu-Mao Li
Efficiently reuses Monte Carlo gradient samples between optimization steps for differentiable and inverse rendering.

Open Source Projects

Javelin
Just-in-time shader programming embedded in C++ using metaprogramming and operator overloading techniques.

Howler
Light-weight logging and error reporting system with a structured and configurable format. Integrates naturally into existing debuggers.

Nabu
Parser combinator framework for C++ that interoperates with user-defined types using metaprogramming and memoization for effiency.

Kobra
Research-oriented rendering engine written with C++ and Vulkan, with implementations for modern algorithms like ReSTIR.

littlevk
Lightweight, single-header C++ wrapper for Vulkan to reduce boilerplate code and accelerate graphics application development.

Zhetapi
Machine learning and numerical analysis library written in C++ for scientific computing. Includes a user friendly Python-like scripting language.

SMake
Simple YAML-based build system for C/C++ projects. Provides automatic dependency detection, and cross-platform compilation support.

metacpp
Experiments on the limits of C++ metaprogramming. Benchmarks multiple implementations of a compile-time Lisp evaluator.

tuicpp
Wrapper library for ncurses that offers intuitive abstractions for common constructs such as text windows, prompts, and tables.

Murphy
Exploratory project for efficiently deforming curves to match a target silhouette using differentiable rasterization.