Call For Papers

We invite authors to submit anonymized papers of up to 4 pages that explore the challenges of aligning multimodal systems with the needs of a pluralistic society. We specifically seek work that addresses the "full stack" of alignment, from shared conceptual foundations and ethical data practices to evaluation metrics that capture cultural nuance. Below, we provide a non-exhaustive list of topics we hope to cover. 

Evaluation, benchmarks, and metrics:

• • Developing new metrics to measure cultural authenticity, pluralism, and fairness across diverse languages, regions, and communities

Data collection and community participation:

• • Methodologies for including local communities in dataset creation and annotation to ensure ethical, representative, and inclusive data

Modeling for Pluralism:

• • Techniques for capturing multiple cultural perspectives and symbolism in models without defaulting to a single normative view

Multilingual & Low-Resource Alignment:

• • Building multimodal models that can amplify underrepresented communities and handle low-resource data at scale

Societal Alignment & Accountability:

• • Investigating the risks of bias, inequity, and cultural erasure, and proposing frameworks for accountability in AI development

Interdisciplinary Perspectives & Shared Vocabulary:

• • Bridging gaps between computer vision, NLP, HCI, and the social sciences to create a shared framework for studying culture in AI

Submission Instructions:

We invite authors to submit anonymized papers of up to 4 pages, excluding references and appendices, in PDF format through the OpenReview submission portal. Submissions must follow the CVPR 2026 template. We welcome works in progress, position papers, policy papers, and academic research papers.

Timeline

• Call for workshop papers: TBD

• Paper submission deadline: TBD

• Notification of acceptance: TBD

• Workshop date: TBD

All accepted papers will be available on the workshop website, but will be considered non-archival.

OpenReview link to submit papers coming soon!

Dual Submission

We encourage submissions of relevant work that has been previously published, is in progress, or is to be presented at the main conference. The accepted papers are non-archival and will not appear in official CVPR proceedings. 

Contact

For any queries please drop an email at maps.cvpr@gmail.com