2025 – Present

I am a co-founder / CEO of Outerport. Our mission is to build AI that can design, engineer and optimize advanced manufacturing processes and industrial equipment.

We do so through an AI process engineering agent that can read engineering diagrams, generate designs, run simulations, and optimize design parameters.We also build AI foundation models trained on physical equipment data that can optimize processes 100x better than existing methods.

We are proud to be working with some of the world's most important companies in chemicals, semiconductor equipment, materials, EPC, and heavy industries.

2020 – 2024

Joined NVIDIA full-time as a Research Scientist in the Hyperscale Graphics Systems Group, focusing on fundamental research on simulation technologies. In particular I focused on how to effectively represent the geometry behind physical objects as data. I published over 10 research works with over 4500 citations (in top venues like CVPR, SIGGRAPH).

Concurrently, I also did a PhD at the University of Toronto, but I went on leave to start a company. Supervised by my amazing advisors, Sanja Fidler and Alec Jacobson.

2016 – 2020

Earned a bachelors from the University of Waterloo with a major in computer science. Took a whole lot of mechatronics courses. Spent my out-of-school hours building autonomous driving vehicles, starting in hardware systems engineering and later led the entire software division.

Did a couple of internships: Tulip Interfaces (which works on software that helps factories run their operations), and NVIDIA doing hardware-software integration for autonomous vehicles. Later did research at NVIDIA and Preferred Networks (in Tokyo!).

2012 – 2016

Started building robots — first in design, then manufacturing / fabrication, and then was the leader driving around town doing procurement, operating CNC mills, building schedules, and whatever needed to be done.

1998 – 2012

Born in Oregon, the land of semiconductor manufacturing. Growing up around Japanese technology and culture was a source of pride as an immigrant child. Influenced by biographies of Edison and Metal Gear Solid, learned the good and bad sides of technology. Learned to program and do 3D modeling. Aspired to invent technologies (especially robots) that make the world a better place.

Publications

FastAtlas: Real-Time Compact Atlases for Texture Space Shading

Eurographics 2025

FastAtlas: Real-Time Compact Atlases for Texture Space Shading

Nicholas Vining, Zander Majercik, Floria Gu, Towaki Takikawa, Ty Trusty, Paul Lalonde, Morgan McGuire, Alla Sheffer

GPU-based atlas packing for real-time texture space rendering.

A LoRA is Worth a Thousand Pictures

arXiv 2024

A LoRA is Worth a Thousand Pictures

Chenxi Liu, Towaki Takikawa, Alec Jacobson

LoRAs > embeddings for retrieval and clustering.

Lagrangian Hashing for Compressed Neural Field Representations

ECCV 2024

Lagrangian Hashing for Compressed Neural Field Representations

Shrisudhan Govindarajan, Zeno Sambugaro, Ahan Shabhanov, Towaki Takikawa, Weiwei Sun, Daniel Rebain, Nicola Conci, Kwang Moo Yi, Andrea Tagliasacchi

Points + grids + hash tables for neural field compression.

What You See is What You GAN: Rendering Every Pixel for High-Fidelity Geometry in 3D GANs

CVPR 2024

What You See is What You GAN: Rendering Every Pixel for High-Fidelity Geometry in 3D GANs

Alex Trevithick, Matthew Chan, Towaki Takikawa, Umar Iqbal, Shalini De Mello, Manmohan Chandraker, Ravi Ramamoorthi, Koki Nagano

Adaptive sampling in rendering = higher fidelity geometry.

Compact Neural Graphics Primitives with Learned Hash Probing

SIGGRAPH Asia 2023

Compact Neural Graphics Primitives with Learned Hash Probing

Towaki Takikawa, Thomas Müller, Merlin Nimier-David, Alex Evans, Sanja Fidler, Alec Jacobson, Alexander Keller

Learned hash probing + hash tables = more compression & faster.

ATT3D: Amortized Text-To-3D Object Synthesis

ICCV 2023

ATT3D: Amortized Text-To-3D Object Synthesis

Jonathan Lorraine, Kevin Xie, Xiaohui Zeng, Chen-Hsuan Lin, Towaki Takikawa, Nicholas Sharp, Tsung-Yi Lin, Ming-Yu Liu, Sanja Fidler, James Lucas

Optimize 3D generation over more prompts -> faster convergence.

Magic3D: High-Resolution Text-to-3D Content Creation

CVPR 2023

⭐ Highlight (Top 2%)

Magic3D: High-Resolution Text-to-3D Content Creation

Chen-Hsuan Lin*, Jun Gao*, Luming Tang*, Towaki Takikawa*, Xiaohui Zeng*, Xun Huang, Karsten Kreis, Sanja Fidler, Ming-Yu Liu, Tsung-Yi Lin

Scaling up 3D generation with smarter data structures.

Variable Bitrate Neural Fields

SIGGRAPH North America 2022

Variable Bitrate Neural Fields

Towaki Takikawa, Alex Evans, Jonathan Tremblay, Thomas Müller, Morgan McGuire, Alec Jacobson, Sanja Fidler

Learned indexing = more compressive than hash tables and grids.

A Dataset and Explorer for 3D Signed Distance Functions

JCGT 2022

A Dataset and Explorer for 3D Signed Distance Functions

Towaki Takikawa, Andrew Glassner, Morgan McGuire

Dataset of shapes made with math equations.

Neural Fields in Visual Computing and Beyond

EG / CGF 2022

Neural Fields in Visual Computing and Beyond

Yiheng Xie, Towaki Takikawa, Shunsuke Saito, Or Litany, Shiqin Yan, Numair Khan, Federico Tombari, James Tompkin, Vincent Sitzmann, Srinath Sridhar

Survey of representing vector fields with neural networks.

RTMV: A Ray-Traced Multi-View Synthetic Dataset for Novel View Synthesis

Tech Report 2022

RTMV: A Ray-Traced Multi-View Synthetic Dataset for Novel View Synthesis

Jonathan Tremblay, Moustafa Meshry, Alex Evans, Jan Kautz, Alexander Keller, Sameh Khamis, Charles Loop, Nathan Morrical, Thomas Müller, Koki Nagano, Towaki Takikawa, Stan Birchfield

Really big dataset of synthetic multi-view images.

Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes

CVPR 2021

⭐ Oral Presentation (Top 4%)

Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes

Towaki Takikawa*, Joey Litalien*, Kangxue Yin, Karsten Kreis, Charles Loop, Derek Nowrouzezahrai, Alec Jacobson, Morgan McGuire, Sanja Fidler

Differentiable data structures + neural fields + GPU programming = fast.

Gated-SCNN: Gated Shape CNNs for Semantic Segmentation

ICCV 2019

Gated-SCNN: Gated Shape CNNs for Semantic Segmentation

Towaki Takikawa*, David Acuna*, Varun Jampani, Sanja Fidler

Pay attention to object boundaries for better computer vision.

Portfolio

NVIDIA Kaolin Core

NVIDIA Kaolin Core

GPU / CUDA kernels for 3D machine learning.

NVIDIA Kaolin Wisp

NVIDIA Kaolin Wisp

GPU-accelerated library for neural fields.

WATonomous

WATonomous

Built an autonomous vehicles from scratch.

Fun Things

  • 2 themes in my life have always been art and machines. Computer graphics is a good medium between them.
  • Outside of art and machines, I enjoy basketball, soccer, football, boxing, violin, and guitar.