Clément Jambon 🚴‍♂️

I'm a first-year PhD student in the Algorithmic Design Group at MIT CSAIL, advised by Prof. Mina Konaković Luković. Before that, I visited the 3D Vision Lab of Seoul National University, 🇰🇷 where I completed my master's thesis under the supervision of Prof. Young Min Kim and Prof. Olga Sorkine-Hornung. Even earlier, I was a research intern at NVIDIA Research under the supervision of Thomas Müller and Merlin Nimier-David.

I'm interested in computer graphics and its intersections with machine learning. Previously, my research was focused on high-performance, generalizable, and manipulable neural rendering and generation techniques. These days, I'm exploring new algorithms and representations to design objects 🪑, robots 🤖, and materials 🧩 that can live in the real world. Some of my recent interests also include computer-aided design, computational fabrication, geometry processing, generative modeling, symmetries of all kinds, and funky Monte Carlo estimators 🖥️.

In my free time, I enjoy hiking ⛰️, cycling up (and down) steep hills 🚴, running with a compass and a map 🗺️, swimming in freezing lakes and rivers 🏊🏻, reading books and research papers 📖, watching international movies you've probably never heard of 🎥, and first and foremost talking and socializing 👋. So if you want to chat or grab a beer, count me in! 🍻

Email / Scholar / CV / Twitter / Bluesky / Github / LinkedIn

Loading...

Education

MIT
Computer Science and Artificial Intelligence Laboratory
PhD program
Supervised by Prof. Mina Konaković Luković in the Algorithmic Design Group
September 2024 - now

ETH Zürich
Department of Computer Science
Master of Science
Major in Visual and Interactive Computing, Minor in Machine Learning
September 2022 - August 2024

École polytechnique
Department of Computer Science
Postgraduate engineering degree
Major in Image, Vision and Learning
September 2019 - August 2023

Publications

BrepDiff: Single-stage B-rep Diffusion Model (SIGGRAPH 2025)
Mingi Lee^*, Dongsu Zhang^*, Clément Jambon^*, Young Min Kim, ^*equal contribution
BrepDiff is a simple, single-stage diffusion model for generating Boundary Representations (B-reps). It generates B-reps by denoising point-based face samples with a dedicated noise schedule. Unlike multi-stage methods, BrepDiff enables intuitive, editable geometry creation, including completion, merging, and interpolation, while achieving competitive performance on unconditional generation.
Coming Soon!

NeRFshop: Interactive Editing of Neural Radiance Fields (I3D 2023)
Clément Jambon, Bernhard Kerbl, Georgios Kopanas, Stavros Diolatzis, Thomas Leimkühler, George Drettakis
NeRFshop is an end-to-end method that allows users to interactively select and deform objects through cage-based transformations within NeRFs. Once complete, edits can be collapsed and saved as a portable NeRF representation through a distillation process.
Project page | Paper | Video | Talk | Slides

Neural Point Catacaustics for Novel-View Synthesis of Reflections (SIGGRAPH Asia 2022)
Georgios Kopanas, Thomas Leimkühler, Gilles Reiner, Clément Jambon, George Drettakis
Neural Point Catacaustics is a new point-based representation that allows novel-view synthesis of scenes with curved reflectors, from a set of casually-captured input.
Project page | Paper | Video

Experience

3D Vision Lab
Visiting Research Student
Supervisor: Young Min Kim
February 2024 - August 2024

NVIDIA
Research Intern
Supervisors: Thomas Müller and Merlin Nimier-David
October 2023 - February 2024

Inria - GraphDeco Team
Research Intern
Supervisor: George Drettakis
March 2022 - August 2022

Wemap
Computer Vision Engineer (Internship)
Supervisor: Thibaud Michel
June 2021 - August 2021

Institute of Technology of Cambodia
Teacher and Research Assistant (Internship)
October 2019 - March 2020

Projects

Improved 3D scene reconstruction with diffusion models (ETH Zürich)
Semester project supervised by Silvan Weder among the Computer Vision and Geometry Lab
During 5 months, I investigated large-scale 2D diffusion models in order to perform "neural extrapolations" of partial NeRF scenes. To tackle this problem, I mostly focused on the following state-of-the-art techniques: Feature Fields, Score Distillation Sampling (SDS) and feature-guided conditioning of diffusion models.
Report (71.7 MB) | Low-res report (9.7 MB)

Game Theory for Image Segmentation (ETH Zürich)
For the block course Controversies in Game Theory IX: Cooperative and Non-Cooperative Game Theory, I reproduced results from two previous works and implemented from scratch a segmentation algorithm based on game theory and evolutionary dynamics. Additionally, I investigated the use of DINO features for semantic segmentation.
Code | Report

Computer Graphics 2022: Rendering Competition (ETH Zürich)
Honorable Mention of the jury
For the Computer Graphics 2022 Rendering Competition at ETH Zürich, Marius Debussche and I implemented a path tracer with complex features including spectral rendering, subsurface scattering, an advanced camera model and participating media.