Regular papers
Full papers following ECCV format, published in workshop proceedings (14 pages).
Computer Vision/CORE Unranked
UNCV 2026
5th Workshop on Uncertainty Quantification for Computer Vision
The 5th Workshop on Uncertainty Quantification for Computer Vision (UNCV 2026) is an ECCV 2026 affiliated workshop aiming to advance the integration of predictive uncertainty into computer vision systems, particularly in light of Generative AI. It brings together researchers to discuss methods for reliable, well-calibrated vision models that can recognize their own limitations, with a focus on safety-critical applications, robustness, and multimodal systems.
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Extended abstracts
Short papers up to 4 pages, not published in proceedings.
Papers accepted at ECCV 2024
Submissions of papers already accepted at ECCV 2024, submitted in their original accepted format.
Research areas in scope
01
Applications of Uncertainty Quantification
Applications of uncertainty quantification in computer visionReliability of Multi-modal Models (e.g., Vision-Language)Uncertainty Quantification of Open-Vocabulary approachesFailure prediction (e.g., Out-of-Distribution detection)Robustness in computer visionSafety critical applications (e.g., autonomous driving)Domain-shift in computer visionUncertainty Quantification of GenAI approaches
02
Methods for Uncertainty Quantification
Deep probabilistic modelsMethods for uncertainty quantificationIncorporating explicit prior knowledge in deep learningOutput ambiguity, label noise, and diversity
Policies worth checking twice
- All submissions will be peer-reviewed.
- Only regular ECCV workshop papers will be included in the Workshop Proceedings.
- Submissions must follow ECCV author guidelines for regular papers.
- Extended abstracts must use the UNCV 2024 style (based on CVPR 2022 style) and be up to 4 pages, excluding references.
- Accepted ECCV 2024 papers must be submitted in their original accepted format.
- Supplementary material may be uploaded separately on OpenReview, but reviewers are not required to consult it.
- Submissions are handled via OpenReview.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Official workshop siteCall for PapersAccepted PapersProgram2nd Workshop on Uncertainty Quantification for Computer Vision at ICCV 20233rd Workshop on Uncertainty Quantification for Computer Vision at ECCV 2024
Last verified September 9, 2026