Full Papers
Eight-page papers (including figures and tables) using the KnowledgeMR Workshop LaTeX template, with additional pages allowed for references and appendix. Papers must be anonymized for double-blind review.
Computer Vision/CORE Unranked
KnowledgeMR 2026
2nd Workshop on Knowledge-Intensive Multimodal Reasoning
Jun 12 2025Official workshop site Site reachable
KnowledgeMR 2026 is a workshop co-located with CVPR 2026 that focuses on advancing multimodal AI systems capable of reliable, transparent, and contextually grounded reasoning in specialized, high-stakes domains such as medicine, engineering, and scientific research. The workshop brings together researchers to share innovations in knowledge-intensive multimodal learning, foundation models, embodied AI, evaluation benchmarks, and ethical considerations through invited talks, panels, and poster sessions.
Key deadlines
Verified September 9, 2026Full paper
April 17, 2026
AoE
Workshop timeline
Submission and decisions
Full paperKey deadline
April 17, 2026 · AoE
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Research areas in scope
01
Knowledge-intensive Multimodal Learning
Method and architecture designs integrating domain-specific knowledge with diverse data sources (e.g., text, images, sensor data, structured data)Data curation strategiesModality fusion techniquesRepresentation learning frameworksExplainability methods for reliable reasoning in high-stakes settings
02
Multimodal Foundation Models for Specialized Domains
Adapting large-scale multimodal foundation models for clinical diagnostics, scientific research, and advanced engineeringEfficient fine-tuning strategiesPrompt engineering for domain expertiseDomain-centric pre-trainingKnowledge distillation to blend foundational capabilities with expert-level insights
03
Embodied AI for Knowledge-Intensive Scenarios
Integration of multimodal reasoning in physical or interactive domains (e.g., industrial automation, laboratory robotics)Sensor fusionAdaptive learning with minimal supervisionHuman-robot collaborationSimulation-to-real transfer in safety-critical scenariosEnsuring transparency, robustness, and trustworthiness in embodied AI systems
04
Evaluation and Benchmarking
Development of standardized benchmarks for domain-specific multimodal reasoningPerformance metrics tailored to specialized domain challengesTesting methodologies capturing the full spectrum of domain-specific difficulties
05
Broader Topics in Knowledge-Intensive Multimodal Reasoning
Integrating symbolic and neural methods for structured reasoningEnsuring privacy and security with sensitive dataExploring multi-agent collaboration for complex decision-makingSocietal and ethical considerations in deploying multimodal systems in real-world, high-stakes environments
Policies worth checking twice
- Submissions are double-blind: authors and reviewers must remain anonymous; failure to anonymize results in desk rejection.
- Papers are limited to eight pages (excluding references and appendix); non-compliance leads to rejection without review.
- Submissions under review at other venues are permitted, provided they do not violate those venues' dual-submission or anonymity policies.
- Accepted papers will not be indexed or have archival proceedings.
- Authors must maintain up-to-date OpenReview profiles; incomplete profiles may result in desk rejection.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026