Christoph Lassner

Christoph Lassner is a German computer scientist working at the intersection of machine learning, computer vision, and computer graphics, and a co-founder of World Labs, the spatial-intelligence company launched in 2024 by Fei-Fei Li.
Nextomoro
8 min read
Christoph Lassner

Christoph Lassner

Christoph Lassner is a German computer scientist whose research spans machine learning, computer vision, and computer graphics, with a focus on reconstructing and rendering humans and 3D scenes from photographs and video. He is a co-author of the 2016 Keep it SMPL paper that introduced SMPLify, the first method to automatically estimate 3D human pose and shape from a single image, and lead author of Pulsar, the differentiable sphere-based renderer that became the sphere-rendering backend of PyTorch3D. As of May 2026, he is a co-founder of World Labs, the San Francisco spatial-intelligence company launched in 2024 by Fei-Fei Li, having joined from research-leadership roles at Epic Games and Meta Reality Labs Research.

At a glance

Origins

Lassner is German and grew up in Augsburg, Bavaria. He completed his Diplom in computer science at the University of Augsburg in 2012 with a diploma thesis titled "An Analysis of Successful Approaches to Human Pose Estimation" that established the research line he would carry through the PhD and into industry, and served as a scientific staff member at the Augsburg machine-learning-and-vision chair before moving to Tübingen for doctoral study.

Lassner entered the Max Planck Institute for Intelligent Systems in Tübingen as a doctoral candidate in the Perceiving Systems department, advised by Peter Gehler, with affiliation to the Bernstein Center for Computational Neuroscience at the University of Tübingen. The PhD completed in July 2017, with the dissertation focused on the recognition and detailed 3D reconstruction of human bodies from images.

Career

The Tübingen doctoral years from approximately 2012 to 2017 produced Lassner's foundational publication record on human-body modeling, with Gehler as advisor and a collaborator network around Michael J. Black (the Perceiving Systems department director and co-creator of the SMPL body model), Federica Bogo, Angjoo Kanazawa, and Javier Romero. The 2016 ECCV paper "Keep it SMPL" introduced SMPLify, the first automatic system to fit the SMPL body model to a single image from 2D joint detections; it has accumulated more than 2,200 Google Scholar citations and seeded the model-based human-pose-and-shape line that SPIN, HMR, and SMPLify-X built on.

In September 2017 Lassner joined Body Labs, a New York body-modeling startup co-founded by Black with technology rooted in the Tübingen lab. Amazon acquired Body Labs in October 2017 for an estimated $50 million to $70 million, and Lassner continued at Amazon as a research-and-engineering lead on the body-modeling pipeline that became the core of the Amazon Halo wellness band's body-scan feature.

In March 2019 Lassner joined Facebook Reality Labs Research (renamed Meta Reality Labs Research in October 2021) as a research scientist, initially in Sausalito and later associated with the Pittsburgh laboratory hosting the company's photoreal-avatar research. The Meta years produced Lassner's principal neural-rendering publication record alongside Michael Zollhöfer, Tony Tung, Edgar Tretschk, and Stephen Lombardi. The 2021 CVPR paper "Pulsar: Efficient Sphere-based Neural Rendering" introduced a differentiable renderer representing scenes as collections of spheres with a parallel projection operation; Pulsar was integrated as the sphere-based rendering backend of PyTorch3D and is widely viewed as a precursor to 3D Gaussian splatting.

In approximately 2022 Lassner moved to Epic Games as a research lead, working on real-time machine-learning techniques for digital-human creation in the company's MetaHuman and Unreal Engine ecosystem; the tenure ran to early 2024.

In early 2024 Lassner co-founded World Labs with Fei-Fei Li (chief executive officer), Justin Johnson (a former Stanford PhD student of Li's, on partial leave from the University of Michigan), and Ben Mildenhall (formerly Google Research, lead author of NeRF). The founding thesis is that "spatial intelligence," AI systems that understand, generate, and reason about three-dimensional environments, is a research direction separable from the dominant LLM and 2D-image-generation paradigms. World Labs raised a $230 million seed at a $1 billion-class valuation in September 2024 (Andreessen Horowitz, Radical Ventures, NEA leads) and a $1 billion strategic round in February 2026, with Autodesk contributing $200 million as strategic anchor. The first commercial product, Marble, generates and edits persistent 3D environments from text, images, video, or 3D layouts.

Affiliations

Notable contributions

  • Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image (ECCV 2016). With Federica Bogo, Angjoo Kanazawa, Peter Gehler, Javier Romero, and Michael J. Black. Introduced SMPLify, the first automatic system to fit the SMPL body model to a single image from 2D joint detections. The most-cited single paper in Lassner's record (more than 2,200 Google Scholar citations) and a foundational reference for model-based human-pose-and-shape estimation.
  • Unite the People: Closing the Loop Between 3D and 2D Human Representations (CVPR 2017). With Romero, Martin Kiefel, Bogo, Black, and Gehler. First-author Tübingen-era paper introducing the UP-3D dataset and a unified 2D-to-3D human-representation framework.
  • A Generative Model of People in Clothing (ICCV 2017). With Gerard Pons-Moll and Gehler. First-author paper introducing ClothNet, an early generative image model for full-body clothed humans.
  • Neural Body Fitting (3DV 2018). With Mohamed Omran, Gehler, and Bernt Schiele. Combination of deep learning and the SMPL body model for image-based 3D human-body reconstruction.
  • Pulsar: Efficient Sphere-based Neural Rendering (CVPR 2021). With Michael Zollhöfer. First-author paper introducing the Pulsar differentiable renderer, a sphere-based scene representation enabling real-time rendering of scenes with millions of spheres. Integrated into PyTorch3D and widely regarded as a precursor to 3D Gaussian splatting.
  • Non-Rigid Neural Radiance Fields (ICCV 2021). With Edgar Tretschk, Ayush Tewari, Vladislav Golyanik, Zollhöfer, and Christian Theobalt. Novel-view synthesis of dynamic, non-rigidly deforming scenes from monocular video.
  • Neural 3D Video Synthesis from Multi-View Video (CVPR 2022). With Tianye Li, Mira Slavcheva, Zollhöfer, Simon Green, and others. Time-conditioned neural radiance fields for dynamic-scene capture from a sparse multi-view rig.
  • ARCH: Animatable Reconstruction of Clothed Humans (CVPR 2020). With Zeng Huang, Yuanlu Xu, Tony Tung, and Hao Li. End-to-end framework for reconstructing animatable 3D clothed-human models from a single image.
  • Advances in Neural Rendering (Eurographics 2022 state-of-the-art report). With Tewari, Justus Thies, Ben Mildenhall, Pratul Srinivasan, Tretschk, Zollhöfer, Golyanik, and others. Review paper synthesizing neural rendering through 2021.
  • World Labs co-founding and the spatial-intelligence research program (2024 onward). Marble is the first commercial product.

Investments and boards

The entries below are limited to AI, semiconductors, datacenters, software, and energy.

  • World Labs (AI): Co-founder, 2024 to present. $230 million seed at a $1 billion-class valuation, September 2024 (Andreessen Horowitz, Radical Ventures, NEA leads); $1 billion strategic round in February 2026 with Autodesk contributing $200 million as strategic anchor.

No personal angel investments in AI, semiconductors, datacenters, software, or energy companies have been publicly disclosed as of May 2026.

Network

Lassner's longest-running professional relationship is with Peter Gehler, his Tübingen PhD advisor and co-author on the SMPLify, Unite the People, ClothNet, and Neural Body Fitting papers. The Tübingen Perceiving Systems community around Michael J. Black, the department director and senior co-author on SMPLify and Unite the People, included Federica Bogo, Javier Romero, Angjoo Kanazawa, and Gerard Pons-Moll. Black, co-creator of SMPL and founder of Body Labs, is the figure most directly responsible for the transition that carried Lassner from Tübingen to Body Labs and onward to Amazon.

The Meta Reality Labs Research period produced a separate collaborator network around Michael Zollhöfer, Lassner's most frequent Meta-era co-author across the Pulsar, NR-NeRF, and Neural 3D Video papers. Edgar Tretschk, Christian Theobalt, Tony Tung, and Stephen Lombardi were additional collaborators; the Eurographics 2022 Advances in Neural Rendering state-of-the-art report connected Lassner directly to the broader NeRF community including Ben Mildenhall and Pratul Srinivasan.

The World Labs co-founder team is Lassner's primary current commercial collaborator group: Fei-Fei Li, chief executive officer; Justin Johnson, University of Michigan computer-vision faculty; and Ben Mildenhall, lead author of NeRF.

Position in the field

Lassner's standing in the senior computer-vision-and-graphics cohort rests on two distinct bodies of work. The 2016 SMPLify paper anchors the model-based human-pose-and-shape line that has been productive since the mid-2010s; the 2021 Pulsar paper anchors a smaller but technically influential thread in differentiable rendering, cited as a precursor to the 3D Gaussian splatting line that has dominated post-2023 neural rendering. His Google Scholar citation count of approximately 8,300 with an h-index of 23 places him in the senior practitioner band of the neural-rendering and human-reconstruction subfields.

The career trajectory through Body Labs, Amazon, Meta Reality Labs Research, and Epic Games gave Lassner an unusually broad industrial footprint: a startup acquisition, a wellness-product deployment at Amazon scale, a multi-year tenure at the largest industry neural-rendering lab, and a real-time game-engine research role at Epic. The World Labs role frames his current posture as a working co-founder of a commercial frontier-research lab whose thesis continues his published research.

Outlook

Open questions over the next 6 to 18 months:

  • World Labs research and product cadence. Whether Lassner's publication output continues at the Meta-era cadence, and whether World Labs follows Marble with successor products that draw on the human-reconstruction and Pulsar lines.
  • Marble adoption. Marble's traction in gaming, visual-effects, VR, and 3D-design, and whether spatial intelligence becomes a separable market large enough to support a $2 billion-class company.
  • Sphere-based and Gaussian-splatting lineage. Whether World Labs produces follow-on architectural innovations in the rendering line that Pulsar helped seed.
  • Open-source releases. Whether Lassner continues the Pulsar/PyTorch3D open-source pattern alongside Marble.
  • Public profile. Continued conference talks at CVPR, ECCV, ICCV, and SIGGRAPH, where Lassner has been a recurring invited speaker on the neural-rendering research line.

Sources

Person World Labs
About the author
Nextomoro

Nextomoro

nextomoro tracks progress for AI research labs, models, and what's next.

Read next

Ekin Dogus Cubuk

Andrew White

Daniel Balsam

Laurent Massoulié

Patrick Pérez

AI Research Lab Intelligence

nextomoro tracks progress for AI research labs, models, and what's next.

AI Research Lab Intelligence

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to AI Research Lab Intelligence.

Your link has expired.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.