Archival Track (Full Papers)
Original, unpublished work; 6–8 pages main text (excluding references and appendices); included in official CVPR 2026 workshop proceedings.
Computer Vision/CORE Unranked
CV4Edu 2026
Computer Vision × Education: Building a Cross-Community Agenda for Multimodal Vision in Classrooms (Non-Archival Track)
CV4Edu 2026 is a CVPR workshop focused on advancing computer vision for real-world educational settings by bridging computer vision, natural language processing, human-computer interaction, and education research. It aims to develop efficient, privacy-aware multimodal models that generalize to noisy, low-resource classroom environments, with goals including shared datasets, metrics, and unified practices for educational AI.
Key deadlines
Verified September 9, 2026Full paper
April 10, 2026
AoE
Conference timeline
Submission and decisions
Full paperKey deadline
April 10, 2026 · AoE
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Non-Archival Track (Extended Abstracts + Short / Position Papers)
Ongoing/early-stage work, datasets in progress, negative results, lessons learned, position papers, or previously published work; extended abstracts up to 2 pages, short/position papers up to 4 pages (excluding references); not included in official proceedings.
Research areas in scope
01
Multimodal Classroom Perception
Face, gaze, pose, gesture, posture, affect, and prosodyVideo, audio, gaze sensors, and wearables (egocentric and exocentric)Multimodal fusion, representation learning, and cross-view / multi-camera setups
02
Language-Centered Multimodal Learning Analytics
Linking speech/text to video events, gaze/attention, and instructional contextClassroom NLP: ASR robustness, diarization, evaluating and mitigating bias, discourse modeling, dialogue/tutoring interactions, simplification, misconception detectionRetrieval-augmented classroom analytics, model adaptation, evaluation for learning-aligned outcomes
03
Robustness & Generalization
Domain shift beyond the lab, occlusions, noisy data, and missing modalitiesFew-/low-shot learning, continual and on-device adaptationGeneralization across classroom layouts and populations
04
Human Behavior Modeling for Learning
Engagement, attention, affect, confusion, self-regulation, and metacognitionCollaboration, group dynamics, and teacher–student interactionsGaze-informed models, saliency/scanpath prediction, activity recognition
05
Temporal Modeling & Intervention
Sequential/temporal models of learning processesBehavioral forecasting, early-warning systems, and interventionsReal-time inference, feedback, and human-in-the-loop systems
06
Interpretability, Reliability & Evaluation
Interpretable models, uncertainty estimation, and calibrationOOD detection, fairness, and bias analysisEvaluation protocols aligned with learning outcomes
07
Privacy-Aware AI, Datasets & Deployments
Privacy-preserving data collection, anonymization, de-identification, and governanceAnnotation strategies, construct-aligned labeling, active learning, synthetic data, and dataset curationClassroom-ready systems, scalable multimodal data-collection frameworks, edge/on-device inference, and real-world deployments
Policies worth checking twice
- All submissions must follow the CVPR 2026 paper template and official style guidelines
- Each submission undergoes rigorous double-blind review by at least three reviewers
- Conflicts of interest are managed through the OpenReview platform
- Submissions must comply with CVPR policies
- An ethics/IRB checklist is required where applicable
- An optional ethics and broader impact statement may be included
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026