📝 Track Overview
Purpose & Scope
The Extended Abstracts track provides a venue for early-stage, ongoing, or exploratory research that may not yet be ready for full publication. This non-archival format allows you to share exciting ideas, preliminary findings, or valuable negative results while retaining the ability to publish the full work elsewhere.
💡 Why Submit an Extended Abstract?
- Get Early Feedback: Receive expert input on work-in-progress to guide future research
- Share Preliminary Results: Present early findings to the community and spark discussions
- Report Negative Results: Help the field learn from what doesn't work
- Test New Ideas: Explore novel concepts before committing to full-scale experiments
- Network & Collaborate: Connect with researchers working on similar problems
- No Publication Restrictions: Submit the full version to top conferences later
What We Accept
Preliminary Research
Early-stage work with promising initial results but incomplete validation. Perhaps you've tested on a limited dataset or need broader evaluation.
Example: A novel self-supervised learning approach tested on 1-2 datasets showing preliminary improvements.
Work-in-Progress
Ongoing research with partial results or experiments still in progress. Share your direction and get community feedback.
Example: A new medical vision-language model architecture with early promising results, with full evaluation underway.
Novel Perspectives
Fresh viewpoints on existing problems, new problem formulations, or critiques of current approaches backed by initial evidence.
Example: An analysis of why current domain adaptation methods fail in specific medical imaging scenarios.
Negative Results
Well-executed experiments that didn't yield expected outcomes. These are valuable for preventing others from pursuing dead ends.
Example: "Why Foundation Model X Fails for Radiology: A Systematic Study" showing careful analysis of failures.
Exploratory Studies
Investigations of new directions, feasibility studies, or pilot experiments exploring novel clinical applications.
Example: Feasibility study of applying federated learning across 3 hospitals for a specific imaging task.
Dataset & Benchmark
New datasets, benchmarks, or evaluation protocols that are still being validated or collecting data.
Example: A new benchmark for fairness evaluation in medical imaging AI across demographic groups.
Submission Requirements
Format & Length
- Minimum 2 pages, maximum 4 pages (excluding references)
- Use the WACV 2026 template (same as regular papers)
- References do not count toward page limit
- Supplementary material is optional
🎯 Focus on Substance Over Polish
Extended abstracts are not expected to have the complete experimental validation of full papers. Focus on clearly presenting your idea, initial results, and future directions. Reviewers understand this is work-in-progress.
Anonymization Requirements
Extended abstracts also undergo double-blind review. Please:
- Remove author names and affiliations
- Anonymize references to your own work
- Remove acknowledgments
- Avoid identifying information in text or figures
Suggested Structure
Your extended abstract should include:
1. Title & Abstract (200 words)
Clear, concise summary of your work and findings
2. Introduction (0.5-1 page)
- Clinical or technical motivation
- What gap are you addressing?
- Your main idea or hypothesis
3. Approach (0.5-1 page)
- Brief description of your method
- Key technical details
- What makes it novel or interesting?
4. Preliminary Results (0.5-1 page)
- Dataset(s) used
- Initial experimental findings
- Tables or figures showing results
5. Discussion & Future Work (0.5-1 page)
- What do results suggest?
- Current limitations
- Planned next steps
- Open questions for community input
Special Considerations
For Negative Results
When submitting negative results, please ensure:
- Rigorous Methodology: Experiments were properly designed and executed
- Clear Hypotheses: What did you expect and why?
- Thorough Analysis: Why didn't it work? What did you learn?
- Actionable Insights: What should others avoid or investigate?
- Reproducibility: Provide enough detail so others can verify findings
✨ We Value Negative Results!
Negative results are scientifically valuable and help the community avoid unproductive directions. A well-documented negative result can be just as important as a positive one. We explicitly encourage these submissions and reviewers will evaluate them on methodological rigor and insights provided, not on whether results are "positive."
For Preliminary Work
If your work is still early-stage:
- Be Transparent: Clearly state what is preliminary and what is ongoing
- Show Promise: Include initial results that suggest the approach is worth pursuing
- Seek Feedback: Identify specific aspects where you'd like community input
- Discuss Plans: Outline what experiments you plan to complete the work
Review Process
Submission
Submit via portal by November 28, 2025
Light Desk Review
Quick check for topic relevance and basic requirements
Peer Review
2-3 reviewers assess quality, novelty, and potential impact
Notification
Decision by January 2, 2026 with constructive feedback
Review Criteria (More Lenient than Regular Papers)
Is the idea or approach interesting and worth discussing?
Is the approach reasonable (even if incomplete)?
Is the abstract well-written and easy to understand?
Could this lead to important findings?
Will this spark useful discussions at the workshop?
Does it provide valuable insights (even if results are negative)?
Note: Reviewers understand these are preliminary works. The bar is for "promising and interesting" rather than "complete and polished."
Presentation & Visibility
Accepted Extended Abstracts Will:
- ✓ Be presented as posters during dedicated poster sessions
- ✓ Be listed on the workshop website with PDFs available
- ✓ Receive feedback from expert reviewers to guide future work
- ✓ Have opportunities to discuss with keynote speakers and attendees
- ✗ Will NOT be published in IEEE/CVF proceedings (non-archival)
✅ Publication Freedom
Because extended abstracts are non-archival, you are FREE to:
- Submit the full version to CVPR, ICCV, ECCV, MICCAI, or any other conference
- Publish the complete work in journals
- Include this work in your thesis
- Present similar work at other workshops
This track lets you share ideas early without sacrificing future publication venues!
Poster Presentation
All accepted extended abstracts will present posters during the workshop:
- Format: Standard poster size (details will be provided upon acceptance)
- Duration: 90-minute poster session with coffee
- Interaction: Opportunity for in-depth discussions with attendees
- Virtual Option: For remote presenters (if available)
Tips for a Strong Extended Abstract
🎯 Clear Motivation
Even if preliminary, clearly articulate WHY this problem matters and why your approach is interesting.
🔬 Show What You Have
Don't wait for perfection. Show your preliminary results, even if limited. Reviewers want to see potential.
💭 Be Honest
Acknowledge limitations and what's incomplete. Reviewers appreciate transparency about work-in-progress.
❓ Ask Questions
Include specific questions where you'd like community input. This makes for better workshop discussions.
📈 Show Promise
Include at least some evidence that your approach is worth pursuing (even if just on toy data).
🗺️ Future Roadmap
Clearly outline your plans to complete the work. This shows you're serious about the research.
Example Scenarios
Scenario 1: PhD Student with Early Results
Situation: You've developed a new active learning strategy for medical image segmentation and tested it on 2 datasets with promising results, but need more extensive evaluation.
What to do: Submit as extended abstract! Present your approach, show preliminary results on the 2 datasets, acknowledge you need broader validation, and ask for feedback on experimental design.
Scenario 2: Negative Result from Industry
Situation: Your team tried applying a popular foundation model to your clinical task and it failed spectacularly. You've analyzed why it failed.
What to do: Perfect for extended abstract! Document what you tried, show the failure modes, analyze why it didn't work, and save others from the same mistake.
Scenario 3: Novel Idea, Limited Resources
Situation: You have a creative idea for multi-modal learning but limited computational resources. You've tested on a small-scale version and it works.
What to do: Submit as extended abstract! Present the idea, show proof-of-concept results, and discuss what large-scale experiments you'd like to do. You might even find collaborators!
Important Dates
Frequently Asked Questions
Q: Can I submit the same work as both extended abstract and regular paper?
A: No. Choose one track per submission. If you have complete results, submit as regular paper. If preliminary, choose extended abstract.
Q: If my extended abstract is accepted, can I still submit to CVPR/ICCV?
A: Yes! Extended abstracts are non-archival. You're free to submit the full version anywhere.
Q: How much results do I need for an extended abstract?
A: You need at least some preliminary evidence. Even results on 1-2 datasets or limited experiments are fine, as long as they show promise.
Q: Are negative results really acceptable?
A: Absolutely! Negative results with good analysis are valuable. We explicitly encourage them.
Q: Can I cite my extended abstract in future papers?
A: Yes, you can cite it as a workshop extended abstract (though it won't be in IEEE proceedings).
Q: What if my work is rejected as extended abstract?
A: Use the reviewer feedback to improve and submit elsewhere. The reviews are meant to be constructive.
Questions?
For questions about the Extended Abstracts track, contact us at p2pcv.wacv@gmail.com with "Extended Abstract" in the subject line.
Share Your Early-Stage Research!
Extended abstracts are perfect for getting feedback and starting discussions.