CVPR 2026 Workshop on
Computer Vision for Children (CV4CHL)
Bringing pediatricians, psychologists, educationists, clinicians and CV researchers together to think about the future.
About CV4CHL
Current computer vision (CV) research and applications often prioritize adult-focused solutions, while progress in CV designed specifically for children's development, health, and education has lagged behind. Our workshop aims to spotlight this critical gap and bring together researchers from diverse fields to discuss the future of CV design and its applications for children.
In the era of advanced visual intelligence, developing bespoke CV systems for children holds special significance for healthcare, education, and developmental support.
Where
Denver, Colorado, USA
When
Date: June 4, 2026 | Time: 8:00 a.m. to 12:10 p.m. | Location: Exhibit Hall A 106
Overview
Key themes and research directions we will explore at CV4CHL.
Why Computer Vision for Children?
-
Emerging Frontier
- Cutting-edge CV technologies, such as vision-language models (VLMs) and large multimodal models, have the potential to support children's development, education, and mental health, opening up a vital new frontier for research.
-
Pediatric Healthcare
- CV in pediatric healthcare is essential, as early diagnosis of childhood diseases through medical imaging and video analysis can lead to timely interventions, improving prognoses and reducing infant mortality rates.
-
Global Impact
- CV can provide valuable tools for children in low-resource countries, helping bridge gaps in education, healthcare, and other developmental supports through scalable and accessible visual solutions.
Topics of Interest
-
Vision for Child Development
- Developmental machine learning and computer vision approaches.
- Vision systems for monitoring and supporting child development milestones.
-
Vision for Pediatric Healthcare
- Medical imaging for childhood diseases and conditions.
- Early diagnosis and screening using computer vision.
- Video analysis for developmental disorder detection.
-
Vision for Education
- Educational tools powered by computer vision.
- Assistive technologies for children with special needs.
-
Vision for Child Psychology
- Emotion recognition and mental health monitoring.
- Social interaction analysis and support systems.
-
Ethics & Social Impact
- Privacy considerations for CV systems involving children.
- Ethical frameworks for deploying CV technologies with children.
- Social robotics and human-robot interaction for children.
Call for Papers
The Workshop on Computer Vision for Children (CV4CHL) at CVPR 2026 invites submissions presenting innovative research in computer vision that addresses key challenges in child education, psychology, pediatrics, child development, social impact, and ethics. We also welcome broader research exploring the intersection of computer vision and children. Submissions may introduce new models, datasets, benchmarks, or applications. Topics of interest include, but are not limited to: developmental machine learning, computer vision for healthcare, social robotics, vision for education, vision for social impact, and vision for good.
Submission Tracks
- Proceeding Track: Submissions to this track will be considered for publication in the official CVPR 2026 workshop proceedings. Papers should be between 6 to 8 pages in length for the main text, formatted according to the CVPR 2026 paper submission guidelines. There is no page limit for references or appendices
- Non-Proceeding Track: We invite submissions of ongoing projects, preliminary results, or work published elsewhere (e.g., on arXiv or at other conferences). Submissions should be limited to 4 pages (as extended abstracts), formatted according to the CVPR 2026 paper submission guidelines. These papers will not be included in the official proceedings.
Review Process
Each submission will undergo a rigorous double-blind review by at least three reviewers. Conflicts of interest will be managed through the OpenReview platform.
Program Committee: A diverse committee of experts from computer vision, pediatrics, psychology, education, and social impact will serve as reviewers. The full list will be announced on the workshop website.
Important Dates (AoE)
- Proceeding Paper Submission Deadline:
March 10, 2026March 15, 2026 - Proceeding Author Notification:
March 17, 2026March 24, 2026 - Proceeding Camera Ready: April 10, 2026
- Non-Proceeding Paper Deadline:
April 09, 2026April 19, 2026 - Non-Proceeding Paper Notification:
April 10, 2026April 22, 2026
Submission Site
Proceeding papers can be submitted through the OpenReview Submission Site.
Non-Proceeding paper and abstract, and Competition can be submitted through the OpenReview Submission Site.
Children Gait Competition
The world's first computer vision challenge for pediatric gait analysis
We are excited to announce the Children Gait Competition as part of our workshop! This pioneering challenge focuses on developing human pose and motion computer vision algorithms for pediatric gait analysis, with applications in automated clinical assessment and disease prediction for children.
Prizes
- 1st Place: Latest AR Glass or Watch
Key Dates
- March 23, 2026: Competition launch & registration opens
- April 23, 2026: Submission deadline
- May 15, 2026: Challenge report submission deadline
- CVPR Workshop: Award ceremony & presentations by top teams
For more details on tasks, dataset access, evaluation metrics, and registration, please visit the Kaggle Competition Page. To submit your report, please use the OpenReview Submission Site.
Final results are out!
The Children Gait Competition winners and full leaderboard are now available on Kaggle.
View the final results here →
Invited Keynote Speakers
Sanmi Koyejo
Stanford University
Jean-Marc Odobez
Idiap Research Institute and EPFL
Dima Damen
University of Bristol and Google DeepMind
James M. Rehg
University of Illinois Urbana-ChampaignEvent Schedule
Date: June 4, 2026 | Time: 8:00 a.m. to 12:10 p.m. | Location: Exhibit Hall A 106
Opening Remark
Xu Cao
Keynote 1
Dr. Sanmi Koyejo
Keynote Title: Do our vision systems actually work for children?
Keynote 2
Dr. Jean-Marc Odobez
Keynote Title: From Gaze Estimation to Child Interaction Understanding
Keynote 3
Dr. Dima Damen
Keynote Title: My World, My View — Learning from an Egocentric Perspective
Panel Discussion
Coffee Chat
Keynote 4
Dr. James M. Rehg
Keynote Title: Quantifying Children's Social Behavior
Competition Paper Talk
Oral Paper Talk
Close Remark
Dr. Ismini Lourentzou
Panelists
Boqing Gong
Boston University
Yu Tian
University of Central Florida
Dima Damen
University of Bristol and Google DeepMind
Yapeng Tian
University of Texas at DallasPanel Discussion Host
Adam Cross MD
University of Illinois and OSF Healthcare
Ismini Lourentzou
University of Illinois Urbana-Champaign
Xu Cao
PediaMed AIOrganizers
Yifan Shen
University of Illinois Urbana-Champaign
Xinzhuo Li
University of Illinois Urbana-Champaign
Boyi Li
University of Illinois Urbana-Champaign
Samy Tafasca
Idiap Research Institute and EPFL
Ismini Lourentzou
University of Illinois Urbana-Champaign
Jintai Chen
HKUST (Guangzhou)
Zhongyi Zhou
Google
Chopard Daphné
University Children's Hospital in Zurich and ETH
Meihuan Huang
Shenzhen Children's Hospital
Yijiang Li
UC San Diego
Ece Özkan Elsen
University of Basel
Guojun Yun
Shenzhen Children's Hospital
Xu Cao
PediaMed AIAdvisory Committee
Jimeng Sun
University of Illinois Urbana-Champaign and Keiji AI
Jean-Marc Odobez
Idiap Research Institute and EPFL
Yinda Zhang
Google
Jianguo Cao
PediaMed AI
Hui Xiong
HKUST (Guangzhou)
James M. Rehg
University of Illinois Urbana-ChampaignSponsors
Platinum Sponsors
Silver Sponsors
Accepted Papers
Highlighted accepted papers for the CV4CHL workshop, divided by proceeding and non-proceeding tracks.
Proceeding
- Annotation-Free Anatomy-Aware Joint Segmentation in Pediatric Hand Radiographs for Explainable Bone Age Estimation
- Unlocking Multi-Site Clinical Data: A Federated Learning Approach to Privacy-First Child Autism Behavior Analysis
- CV4CHL 2026 The First AI Children Challenge
- Age-Inclusive 3D Human Mesh Recovery for Action-Preserving Data Anonymization
- CSA-Graphs: A Privacy-Preserving Structural Dataset for Child Sexual Abuse Research
- Does visual grounding produce more reliable aphasia therapy items than caption-only generation?
- An End-to-End System for 3D Reconstruction of Intraoral Cleft Anatomy in Children from Smartphone Video
- Developmental Geometry in Facial Embeddings: A Boundary-Layer Transition
- A Survey on Embodied AI for Pediatrics
- Child Learning Inspired Multimodal Models: Challenge and Opportunity
Non Proceeding
- BabyVLM-V2: Toward Developmentally Grounded Pretraining and Benchmarking of Vision Foundation Models
- GazeAgent: From Gaze Estimation to Query-Driven Analysis in Naturalistic Interactions
- BabyCLIP: Continual Language Grounding Through the Experience of a Single Child
- All-Age Human Mesh Recovery
- Can AI Understand Indonesian Children's Book Illustrations? Compliance, Safety, and the Cultural Alignment Gap
- Synthetic Data Alone is Enough? Rethinking Data Scarcity in Pediatric Rare Disease Recognition
- A Perspective on Clinical Deployment of AI-GMA: From Quality-Gated Video Acquisition to Robust Decision Support
Contact
Location
CVPR 2026 Venue (Exhibit Hall A 106)
Poster boards: 32-41
GitHub
Email Us
xucao@pediamed.ai