Synthetic Data for Computer Vision @ CVPR 2026

Overview

Welcome to The 3rd Workshop on Synthetic Data for Computer Vision (SynData4CV) at CVPR 2026! During the last decade, advances in computer vision have been catalyzed by the release of meticulously curated human-labeled datasets. Recently, people have increasingly resorted to synthetic data as an alternative to labor-intensive human-labeled datasets for its scalability, customizability, and cost-effectiveness. Synthetic data offers the potential to generate large volumes of diverse and high-quality vision data, tailored to specific scenarios and edge cases that are hard to capture in real-world data. However, challenges such as the domain gap between synthetic and real-world data, potential biases in synthetic generation, and the generalizability of models trained on synthetic data remain. This workshop aims to provide a forum for discussion and encouragement of further exploration in these areas. Topics of interest include, but are not limited to:

Effectiveness: What is the most effective way to generate and leverage synthetic data? Does synthetic data need to "look" realistic?
Efficiency and scalability: Can we make synthetic data generation more efficient and scalable without much sacrifice on the quality?
Benchmark and evaluation: What benchmark and evaluation methods are needed to assess the efficacy of synthetic data for computer vision?
Risks and ethical considerations: How can we mitigate the risks of generating and using synthetic data? How do we address relevant ethical questions, such as bias amplification in synthetic datasets?
Applications: What are other tasks in computer vision or other related fields (e.g., robotics, NLP) that could benefit from synthetic data?
Other open problems: How do we decide which type of data to use, synthetic or real-world data? What is the optimal way to combine both if both are available?

Invited Speakers

Jia Deng

Princeton University

Georgia Gkioxari

Caltech

Nupur Kumari

Carnegie Mellon University

Manling Li

Northwestern University

Andrew Owens

Cornell Tech

Schedule

Date: June 4, 2026 · Afternoon (Half day) Location: Room 607

1:00 – 1:10 PM Opening Remarks Opening
1:10 – 1:45 PM Invited Talk · Manling Li Talk
1:45 – 2:20 PM Invited Talk · Jia Deng Talk
2:20 – 2:55 PM Invited Talk · Georgia Gkioxari Talk
2:55 – 3:10 PM Break Break
3:10 – 3:45 PM Invited Talk · Andrew Owens Talk
3:45 – 4:20 PM Invited Talk · Nupur Kumari Talk
4:20 – 4:30 PM Closing Remarks Closing
4:30 – 5:30 PM Poster Session Poster

Accepted Papers · 54 papers, sorted alphabetically

Addressing Data Scarcity in Depth-Based Human Action Recognition via Zero-Shot Depth Estimation

Rebeka Angyal, Pedro Hermosilla, Martin Kampel, Irene Ballester
AfriST-VQA: Benchmarking MLLMs for Scene-Text Visual Question Answering for African Languages

Henry Gagnier
Appreciate the View: A Task-Aware Evaluation Framework for Novel View Synthesis

Saar Stern, Ido Sobol, Or Litany
Assessing the Predictive Value of Physics-Grounded Synthetic Data for Computer Vision in Space Environments

Arianna Issitt, Emily Happy, Elijah Clark, Mackenzie J. Meni, Ryan T. White
Auto-Comp: Scalable Controlled Synthetic Benchmarks for VL Compositionality

Cristian Sbrolli, Toshihiko Yamasaki, Matteo Matteucci
Avatar4D: Synthesizing Domain-Specific 4D Humans for Real-World Pose Estimation

Jerrin Bright, Zhibo Wang, Dmytro Klepachevskyi, Yuhao Chen, Sirisha Rambhatla, David A. Clausi, John S. Zelek
Beyond Objects: Contextual Synthetic Data Generation for Fine-Grained Classification

William Yang, Xindi Wu, Zhiwei Deng, Esin Tureci, Olga Russakovsky
Beyond Photorealism: Counterfactual Synthetic Bundles for Invariant Sim-to-Real Vision

Murari Ambati
Beyond Raw Signals: Undecoded Generative Latents as Privileged Synthetic Data

Cristian Sbrolli, Nicolas Michel, Matteo Matteucci, Toshihiko Yamasaki
Completing Missing Modalities: Synthetic Data for RGB–Infrared–Thermal–Text Person Re-Identification

Muhammad Umair, Muhammad Hammad Musaddiq, Jun Zhou, Ahmad Muhammad
CryoDiff: Cryo-EM Synthesis via Biophysics and Cycle-Consistent Diffusion

Genpei Zhang, Yuntian Yang, Siqi Wu, Ningyan Zhang, Seonghui Min, Jie Wu, Christopher Braxton Owens, Minhao Wu, Wanyue Feng, Gus LW Hart, Runmin Jiang, Min Xu
Diffusion-Augmented Coreset Expansion for Scalable Dataset Distillation

Ali Abbasi, Shima Imani, Chenyang An, Gayathri Mahalingam, Harsh Shrivastava, Maurice Diesendruck, Hamed Pirsiavash, Pramod Sharma, Soheil Kolouri
Disentangled Anatomy-Disease Diffusion (DADD) for Controllable Ulcerative Colitis Progression Synthesis

Umut Dundar, Alptekin Temizel
Durian: Dual Reference Image-Guided Portrait Animation with Attribute Transfer

Hyunsoo Cha, Byungjun Kim, Hanbyul Joo
Evaluating the Trade-offs of MDL-to-UsdPreviewSurface Material Simplification in NVIDIA Isaac Sim: Visual Quality, Feature Preservation, and AI Task Performance

Zihou Zhu, Mei Haitao, Haolong Zheng, Zhou Zhang
Few-Shot Synthetic Data Generation with Diffusion Models for Downstream Vision Tasks

Daniil Dushenev, Nazariy Karpov, Daniil Zinovjev, Alexander Gorin, Konstantin Kulikov
Fréchet Inception Distance is Failing to Preserve Rank Consistency for Synthetic Out-of-Distribution Samples

Linghui Liu, Henrike Stephani, Janis Keuper
Generating Synthetic Illumination Variation with Co-Located Relighting

Yash Turkar, Karthik K Dantu
Grounding Synthetic Data Generation With Vision and Language Models

Ümit Mert Çağlar, Alptekin Temizel
How Far Can We Go With Synthetic Data for Audio-Visual Sound Source Localization?

Arda Senocak, Sooyoung Park, Tae-Hyun Oh, Joon Son Chung
Hybrid Rendering for Multimodal Autonomous Driving: Merging Neural and Physics-Based Simulation

Máté Tóth, Péter Kovács, Réka Bencses, Zoltan Bendefy, Zoltan Hortsin, Balázs Teréki, Tamas Matuszka
iARCS: Iterative Agentic RL for Controllable 3D Scene Generation

Saugat Adhikari, Ashok Prasad Neupane, Pramish Paudel, Ajad Chhatkuli, Danda Pani Paudel
Masked Language Prompting for Generative Data Augmentation in Few-shot Fashion Style Recognition

Yuki Hirakawa, Ryotaro Shimizu
Multi-Objective Photoreal Simulation (MOPS) Dataset for Computer Vision in Robot Manipulation

Maximilian Xiling Li, Paul Mattes, Nils Blank, Rudolf Lioutikov
Narrowing the Performance Gap in Synthetic VLM Pre-training via Multi-Generator Ensembles

Leonardo Brusini, Cristian Sbrolli, Eugenio Lomurno, Toshihiko Yamasaki, Matteo Matteucci
Object-Centric Data Synthesis for Category-level Object Detection

Vikhyat Agarwal, Jiayi Cora Guo, Declan Hoban, Sissi Yuxi Zhang, Nick Moran, Peter Cho, Srilakshmi Pattabiraman, Shantanu H. Joshi
One Category One Prompt: Dataset Distillation using Diffusion Models

Ali Abbasi, Ashkan Shahbazi, Hamed Pirsiavash, Soheil Kolouri
OrbitArch

I-Ting Tsai, Bharath Hariharan
Personalized Generative Models for Contextual Debiasing

Xinran Liang, Esin Tureci, Prachi Sinha, Ye Zhu, Vikram V. Ramaswamy, Olga Russakovsky
PLLM: Pseudo-Labeling Large Language Models for CAD Program Synthesis

Yuanbo Li, Dule Shu, Yan-Ying Chen, Matthew Klenk, Daniel Ritchie
Privacy-Aware Synthetic Video Benchmarking and Relational Evaluation for Worker-Under-Suspended-Load Detection

Anshu Singh, Alejandro Seif
ProductConsistency: Improving Product Identity Preservation in Instruction-Based Image Editing via SFT and RL

Mukund Khanna, Raj Singh Yadav, Kunal Singh
RareCrafter: Controllable Generative Augmentation for Rare Object Detection in Driving Scenes

Mohadeseh Ghafoori, Danielle Lee, Kurt Hammen, Collin Meese, Mark Nejad
Realiz3D: 3D Generation Made Photorealistic via Domain-Aware Learning

Ido Sobol, Kihyuk Sohn, Yoav Blum, Egor Zakharov, Max Bluvstein, Andrea Vedaldi, Or Litany
Representation-Conditioned Diffusion Models for Guided Training Data Generation

Nithesh Chandher Karthikeyan, Jonas Unger, Gabriel Eilertsen
Restereo: Unifying diffusion stereo video generation and restoration

Xingchang Huang, Ashish Kumar Singh, Florian Dubost, Cristina Nader Vasconcelos, Sakar Khattar, Liang Shi, Christian Theobalt, Cengiz Oztireli, Gurprit Singh
SAIL: Similarity-Aware Guidance and Inter-Caption Augmentation-based Learning for Weakly-Supervised Dense Video Captioning

Ye-Chan Kim, SeungJu Cha, Si-Woo Kim, MinJu Jeon, HyunGee Kim, Dong-Jin Kim
Scaling Up 3D Forest Vision with Synthetic LiDAR

Yihang She, Andrew Blake, David Coomes, Srinivasan Keshav
Sea-Mie: Physically-Based Synthetic Fog for Maritime Image Defogging via Curriculum Learning

Stelios Avlakiotis, Peter Ter Heerdt, Thomas De Kerf, Steve Vanlanduit
Sequential Dataset for Satellite Pose Estimation and a Frequency-Space Neural Operator for HIL-Free Generalization Benchmarking

Woojin Cho, Junghwan Park, Steve Andreas Immanuel, JunminPark, Seokhyun Chin, Jiayun Wang
Sim-to-Real Metrology: Calibrated Digital Twins for Fringe Projection Profilometry

NOBLE AUSTINE, Vuppu Eshwar Sai, Vaishnavi Ravi, Madhu S. Nair, Gorthi Rama Krishna Sai Subrahmanyam
SJEPA: Joint Embedding Predictive Architecture for Synthetic-to-Real Alignment

Shentong Mo
Structure-Consistent Joint Image-Mask Synthesis for Data-Scarce Medical Image Segmentation

Ningyan Zhang, Weiyi Zhang, Mostofa Rafid Uddin, Xingjian Li, Min Xu
Structure-retained low-rank adapters for weather synthesis

Shunxin Wang, Alexandros Stergiou, Luuk Spreeuwers, Estefania Talavera, Nicola Strisciuglio
StyleText: A Large-Scale Dataset and Benchmark for Stylized Scene Text Inpainting

Aleksandr Simonyan, Nipun Jindal
Synthetic Data Generation for Long-Tail Medical Image Classification: A Case Study in Skin Lesions

Jiaxiang Jiang, Mahesh Subedar, Omesh Tickoo
Synthetic Designed Experiments for Diagnosing Vision Model Failures

Krisanu Sarkar
Theory of Space: Evaluating Active Spatial Belief Construction in Foundation Models with Synthetic 3D Environments

Pingyue Zhang, Zihan Huang, Yue Wang, Jieyu Zhang, Letian Xue, Zihan Wang, Qineng Wang, Keshigeyan Chandrasegaran, Ruohan Zhang, Yejin Choi, Ranjay Krishna, Jiajun Wu, Li Fei-Fei, Manling Li
Vanast: Virtual Try-On with Human Image Animation via Synthetic Triplet Supervision

Hyunsoo Cha, Wonjung Woo, Byungjun Kim, Hanbyul Joo
Video-Consistent Synthetic Skiing Trajectories

M'Saydez Campbell, Rémi Emonet, Damien Muselet, Christophe Ducottet
WaterGen: Decoupling Scene and Medium in Underwater Image Generation

Jiayi Wu, Tianfu Wang, Tianyi Xiong, Dehao Yuan, Xiaomin Lin, Md Jahidul Islam, Cornelia Fermuller, Christopher Metzler, Yiannis Aloimonos
When Does Synthetic Data Help? A Spectral Theory of Task-Relevant Domain Gap with Applications to Guided Generation and Bias Auditing

Kaustubh S. Bukkapatnam, Rayan Malik
Why Training with Synthetic Data Fails for OOD: Distribution Gap Amplifies Noise Misalignment in Diffusion Models

Ying Hua, Jessica Bader, Jae Myung Kim, Zeynep Akata
WireSeg-32K: A Physics-Grounded Synthetic Dataset for Wire Instance Segmentation

Zilin Dai, Lehong Wang, Yi Yang, Xiang Fei

Call for Papers

We invite submissions on topics related to synthetic data for computer vision, including but not limited to:

Novel methods for generating synthetic data
Techniques for bridging the domain gap between synthetic and real data
Benchmarks and evaluation metrics for synthetic data
Applications of synthetic data in various computer vision tasks
Ethical considerations and bias mitigation in synthetic data generation
Efficient and scalable synthetic data generation pipelines

Submission Guidelines:

Papers should be formatted according to the CVPR 2026 template
Short papers: 4 pages (excluding references); Long papers: 8 pages (excluding references)
Submissions should be made through OpenReview
All submissions will be double-blind reviewed
Accepted papers will NOT be included in CVPR proceedings, so there are no double submission concerns.