Model Development and Adaptation
Pre-training and fine-tuning LFMs for biomedical and clinical domains
Multimodal data acquisition, curation, and annotation
Development, scalability, and optimization of architectures for biomedical data analysis
Performance Enhancement Techniques
Prompt engineering, chain-of-thought (CoT) reasoning, and retrieval-augmented generation (RAG)
Reinforcement learning with human feedback (RLHF)
Data resampling for class balancing and integration of auxiliary weak learner models
Knowledge Engineering and Representation
Generating knowledge graphs, biological networks, ontologies, and taxonomies
Multimodal alignment for biomedical question answering and decision support
Applications in Biomedicine and Healthcare
Clinical risk prediction, disease diagnosis and prognosis, surgical planning, and personalized treatment
Drug discovery, pharmacogenomics, and personalized medicine
Biomedical imaging applications (radiology, pathology, histology, cells, transcriptomics, etc.)
Visual reasoning and question answering in clinical and biological domains
Nutrition-focused applications (dietary assessment, calorie estimation, meal analysis)
Case studies and real-world deployment in healthcare
Evaluation, Benchmarking, and Reproducibility
Creation of benchmark datasets and evaluation metrics
Reproducibility challenges in biomedical LFM research
Trustworthiness, Ethics, and Societal Impact
Trustworthiness, explainability, hallucination, and validation of results
Equity, fairness, bias, safety, and reliability in biomedical applications
Ethical considerations and interpretability of LFMs
Papers must be limited to eight (8) pages, including figures and tables, following the WACV 2026 formatting style. Additional pages containing only references are permitted. Authors should refer to the WACV 2026 Author Kit for detailed instructions on paper preparation and formatting.
Submissions that are not properly anonymized, do not follow the template, or exceed eight pages (excluding references) will be rejected without review.
All submissions must be made electronically through CMT3. Authors should ensure that their submission is complete, anonymized, and properly formatted before the deadline.
This workshop follows a double-blind review process. Authors’ identities must not be revealed to reviewers and vice versa. To ensure anonymity:
Do not include author names, affiliations, acknowledgments, or grant numbers in the submission.
Avoid references or links that reveal author identity (e.g., institutional repositories, personal websites, identifiable dataset URLs).
Violations of these guidelines may result in desk rejection.
Plagiarism is strictly prohibited. It includes the uncredited use of another’s words, data, figures, or ideas. Any suspected cases will be referred to the IEEE Intellectual Property Office, which may impose sanctions including bans from future IEEE and CVF conferences. You can find information on this office, their procedures, and their definitions of five levels of plagiarism at this webpage. Authors are reminded that plagiarism detection systems are highly accurate and that reviewers often identify reused or copied content.
By submitting to this workshop, authors affirm that:
The work has not been published previously in substantially similar form.
The work is not under review at another journal, conference, or workshop with proceedings.
Submissions with 20% or more overlap with other works will be considered double submissions and rejected.
Preprints such as arXiv or institutional technical reports are not considered prior publications and need not be cited.
If a submission involves personal data or human subjects, the authors must confirm that the data collection and usage comply with relevant ethical standards. Where applicable, authors should indicate whether IRB (Institutional Review Board) or equivalent approval was obtained. If not yet approved, authors should describe the pending status, data source, consent procedures, and any ethical considerations.
Authors using publicly available datasets are encouraged to verify the dataset’s consent and data-use policy.
We follow the NeurIPS 2025 LLM usage guidelines (adapted for this workshop). Authors may use LLMs and related tools in their research or writing, provided that:
The paper clearly describes the use of such tools if they are integral to data processing, analysis, or methodology.
Authors take full responsibility for the accuracy, originality, and integrity of all text and figures produced using such tools.
Any inappropriate use of LLMs for content generation or plagiarism will result in rejection.
At least one author of each accepted paper must register and attend the workshop (either in person or virtually, depending on the conference policy) to present the work. Failure to do so may result in the paper being excluded from the official proceedings.
All accepted papers will be published in the WACV 2026 proceedings (CVF open-access proceedings) and made publicly available approximately two weeks before the main WACV 2026 conference. Authors wishing to submit a patent understand that the paper's official public disclosure is two weeks before the conference or whenever the authors make it publicly available, whichever is first. The conference considers papers confidential until published two weeks before the conference but notes that multiple organizations will have access during the review and production processes, so those seeking patents should discuss filing dates with their IP counsel. The conference assumes no liability for early disclosures. More information about CVF is available at http://www.cv-foundation.org/.
Note to Authors: In cases of uncertainty or conflicting interpretation, authors should refer to the official WACV 2026 Author & Reviewer Guides (available at the CVF/WACV site) for definitive policy clarification.
• Workshop paper submission: Nov 25, 2025 11:59 PM PST
• Workshop paper notification: Dec 24, 2026 11:59 PM PST
• Workshop paper camera-ready: Jan 09, 2026 11:59 PM PST
ACKNOWLEDGEMENT: The Microsoft CMT service was used for managing the peer-reviewing process for this conference. This service was provided for free by Microsoft and they bore all expenses, including costs for Azure cloud services as well as for software development and support.