Short papers
4-page limit plus additional pages for references, formatted in ACM proceedings style, double-blind reviewed, non-archiving, and presented orally at the workshop.
NLP/CORE Unranked
TrustVLM 2026
The 1st Workshop on Toward Trustworthy Vision-Language Models in the Wild: Theory, Algorithm and Application
TrustVLM 2026 is a workshop held in conjunction with ICMR 2026 in Amsterdam, focusing on trustworthy vision-language models (VLMs) in real-world settings. It aims to advance research on theory, algorithms, and applications of trustworthy multi-modal learning, addressing vulnerabilities such as hallucinations, bias, adversarial attacks, and distribution shifts in models like CLIP and LLaVA.
Key deadlines
Verified September 9, 2026Full paper
April 20, 2026
AoE
Workshop timeline
Submission and decisions
Full paperKey deadline
April 20, 2026 · AoE
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Research areas in scope
01
Theory
Uncertainty quantification for VLMsTheoretical bounds on VLMs under distribution shiftsCausal perspectives on spurious correlations and multi-modal shortcut learningRobust optimization and generalization bounds for multi-modal learning objectives
02
Algorithms
Out-of-distribution (OOD) generalization, OOD detection and test-time adaptationDetecting and mitigating hallucination from VLMsRobust training and inference against adversarial attacks, prompt injection and imperfect inputsPost-hoc and intrinsic methods for explaining VLM decisionsTechniques for auditing and debiasing vision-language datasetsMethods to ensure fairness in the context of VLMsPrivacy-preserving machine learning for VLMs
03
Benchmark and Application
Trustworthy usages of VLMs for scientific domains, including program languages, climate science, healthcare, life sciences, physics, and cognitive scienceDetecting AI-generated content and deepfakesNew benchmarks and evaluation protocols for real-world trustworthiness
Policies worth checking twice
- Submissions must be short papers with a 4-page limit (plus references)
- Papers must be formatted according to ACM proceedings style
- Double-blind review process is required
- Submissions must use the specified LaTeX template with \documentclass[sigconf, review, anonymous]{acmart}
- Accepted papers are non-archiving and must be presented in an oral session
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026