Image credit: “Transformer IMU Calibrator: Dynamic On-Body IMU Calibration for Inertial Motion Capture” © Dynamic Calibration for Inertial Motion Capture
The SIGGRAPH Technical Papers program is the premier venue for disseminating and discussing scholarly work in computer graphics and interactive techniques. This program is a core of the conference. At SIGGRAPH 2025, happening 10–14 August in Vancouver, emerging research themes include artificial intelligence, large language models, robotics, and 3D modeling in VR.
SIGGRAPH 2025 Technical Papers saw strong interest with more than 970 submissions — the program’s highest submission pool in all SIGGRAPH’s years. Like in recent years, SIGGRAPH 2025 accepted submissions to two integrated paper tracks: Journal (ACM Transactions on Graphics) and Conference. Additionally, each year, the Technical Papers program honors exceptional research with the Best Papers, Honorable Mentions, and Test-of-Time Awards.
SIGGRAPH 2025 Technical Papers Chair Richard Zhang presents these award-winning papers and thanks the selection committees who chose the Best Papers, Honorable Mentions, and Test-of-Time awardees.
Learn more about the awardees here, and prepare to explore the latest and greatest in computer graphics research at SIGGRAPH 2025.
Best Papers
Shape Space Spectra
We introduce shape-space eigenanalysis to compute eigenfunctions across continuously parameterized shape families. These eigenfunctions are obtained by minimizing a variational principle. To handle eigenvalue dominance swaps at points of multiplicity, we incorporate dynamic reordering during optimization. The method is discretization-agnostic and differentiable, enabling applications in sound synthesis, locomotion, and elastodynamic simulation. Attend the session.
Yue Chang, Otman Benchekroun, Maurizio M. Chiaramonte, Peter Yichen Chen, Eitan Grinspun
CAST: Component-Aligned 3D Scene Reconstruction From an RGB Image
We introduce CAST, an innovative method for reconstructing high-quality 3D scenes from a single RGB image. Supporting open-vocabulary reconstruction, CAST excels in managing occlusions, aligning objects accurately, and ensuring physical consistency with the input, unlocking new possibilities in virtual content creation and robotics. Attend the session.
Kaixin Yao, Longwen Zhang, Xinhao Yan, Yan Zeng, Qixuan Zhang, Jiayuan Gu, Wei Yang, Lan Xu, Jingyi Yu
TokenVerse: Versatile Multi-Concept Personalization in Token Modulation Space
TokenVerse extracts complex visual elements from images by identifying semantic directions in per-token modulation space of DiT models for each word in the image caption. It’s capable of combining concepts from multiple sources by adding corresponding directions, enabling flexible generation of new combinations including abstract concepts like lighting and poses. Attend the session.
Daniel Garibi, Shahar Yadin, Roni Paiss, Omer Tov, Shiran Zada, Ariel Ephrat, Tomer Michaeli, Inbar Mosseri, Tali Dekel
Vector-Valued Monte Carlo Integration Using Ratio Control Variates
Variance reduction techniques are widely used to reduce the noise of Monte Carlo integration. However, these techniques are typically designed with the assumption that the integrand is scalar-valued. To address this, we introduce ratio control variations, an estimator that leverages a ratio-based approach instead of the conventional difference-based control variates. Attend the session.
Haolin Lu, Delio Vicini, Wesley Chang, Tzu-Mao Li
Transformer IMU Calibrator: Dynamic On-Body IMU Calibration for Inertial Motion Capture
This work proposes a dynamic calibration system for inertial motion capture, which can dynamically remove non-static IMU drift and sensor-body offset during usage, enable user-friendly calibration (without T-pose and IMU heading reset), and ensure long-term robustness. Attend the session.
Chengxu Zuo, Jiawei Huang, Xiao Jiang, Yuan Yao, Xiangren Shi, Rui Cao, Xinyu Yi, Feng Xu, Shihui Guo, Yipeng Qin
Honorable Mentions
Lifting the Winding Number: Precise Discontinuities in Neural Fields for Physics Simulation
Yue Chang, Mengfei Liu, Zhecheng Wang, Peter Yichen Chen, Eitan Grinspun
A Monte Carlo Rendering Framework for Simulating Optical Heterodyne Detection
Juhyeon Kim, Craig Benko, Magnus Wrenninge, Ryusuke Villemin, Zeb Barber, Wojciech Jarosz, Adithya Pediredla
Rectangular Surface Parameterization
Etienne Corman, Keenan Crane
High-Performance CPU Cloth Simulation Using Domain-Decomposed Projective Dynamics
Zixuan Lu, Ziheng Liu, Lei Lan, Huamin Wang, Yuko Ishiwaka, Chenfanfu Jiang, Kui Wu, Yin Yang
Variational Surface Reconstruction Using Natural Neighbors
Jianjun Xia, Tao Ju
Moment Bounds Are Differentiable: Efficiently Approximating Measures in Inverse Rendering
Markus Worchel, Marc Alexa
Clebsch Gauge Fluid on Particle Flow Maps
Zhiqi Li, Candong Lin, Duowen Chen, Xinyi Zhou, Shiying Xiong, Bo Zhu
Faraday Cage Estimation of Normals for Point Clouds and Ribbon Sketches
Daniel Scrivener, Daniel Cui, Ellis Coldren, Mazdak Abulnaga, Mikhail Bessmeltsev, Edward Chien
C-Tubes: Design of Tubular Structures From Developable Strips
Klara Mundilova, Michele Vidulis, Quentin Becker, Florin Isvoranu, Mark Pauly
Test-of-Time Awards
ACM SIGGRAPH is proud to announce the 2025 Test-of-Time Award papers that have had a significant and lasting impact on computer graphics and interactive techniques over at least a decade. This is the third year of this annual award. For 2025, the papers presented at SIGGRAPH conferences from 2013 to 2015 were considered by the Test-of-Time Award committee, and the committee selected four winning papers.
Unified Particle Physics for Real-Time Applications (2014)
This paper presents a unified dynamics framework for real-time visual effects by using particles connected by constraints as the fundamental building block. This representation is flexible to model gases, liquids, deformable solids, rigid bodies, and cloth with two-way interactions. It further addresses long-standing computational issues with traditional particle-based methods by introducing a parallel constraint solver based on position-based dynamics. It has enabled numerous real-time applications and has been widely adopted across fields. Read this paper online.
Miles Macklin, Matthias Müller, Nuttapong Chentanez, Tae-Yong Kim
Learning Visual Similarity for Product Design With Convolutional Neural Networks (2015)
This paper was among the first to introduce modern Convolutional Neural Networks (CNNs) to SIGGRAPH, helping to shape the future direction of research in computer graphics and becoming a widely cited reference. In addition to achieving its core goals in visual cross-category and style-aware search across diverse image domains, from complex scenes to isolated objects, it also contributed key technical advancements in CNN training methodologies and in the use of crowdsourcing for collecting training data. Read this paper online.
Sean Bell, Kavita Bala
L1-Medial Skeleton of Point Cloud (2013)
This paper introduces a robust curve skeleton representation for 3D point clouds, which accurately captures shape centers without requiring high-quality input or restrictive geometric or topological constraints. It provides an attractive solution to various applications such as plant, road map, architectures, vessel analysis, and modeling. To this day, it remains a major reference in the field. Read this paper online.
Hui Huang, Shihao Wu, Daniel Cohen-Or, Minglun Gong, Hao Zhang, Guiqing Li, Baoquan Chen
Embree: A Kernel Framework for Efficient CPU Ray Tracing (2014)
This paper introduces the Embree system and describes in detail what it takes to build a professional-grade modular ray tracing framework. It achieved state-of-the-art performance with a simple and elegant architecture and came with an open-source implementation that has been adopted by many since then. The project is still active today and continues to benefit our community. Read this paper online.
Ingo Wald, Sven Woop, Carsten Benthin, Gregory S. Johnson, Manfred Ernst
Register for SIGGRAPH 2025, taking place 10–14 August in Vancouver, to access the leading scholarly research in computer graphics and interactive techniques. View the full schedule to discover all Technical Papers being presented, and be sure to attend the Papers Fast Forward on Sunday, 10 August, at 6 pm PDT.