<aside> 🙌
[Update in Nov, 2024] The leading organizers, Penn State's Ting-Hao 'Kenneth' Huang and C. Lee Giles, are awarded a grant by Alfred P. Sloan Foundation to build and deploy a web-based system that helps writers compose better captions for scientific figures!
</aside>
<aside> 🏆 We're excited to announce the winning teams for this year:
The SciCap organizers and participating teams will present the challenge and their approaches at IJCAI 2024 on August 6. Please come join us!
16:30-16:50 Opening & SciCap (Kenneth Huang)
16:50-17:00 Honorable Mention Team Presentation (Team njust_kmg)
17:00-17:20 Short Caption Track Winner Presentation (Team Allen)
17:20-17:40 Long Caption Track Winner Presentation (Team RB-AI)
17:40-18:00 Discussion (Everyone in the room)
</aside>
<aside> 🔗 Link to this page: SciCap.AI
</aside>
<aside> 🔗 This year, the challenge will be hosted at IJCAI 2024 (August 3-9, Jeju Island, South Korea). The SciCap Challenge will be held on August 6.
</aside>
<aside> 🚨 DEADLINE EXTENSION! We extended the deadline to June 22 (Sat). The technical report deadline is also extended to June 29. Come submit your exciting results!
</aside>
<aside> 🚨 All participating teams should fill out the Google Form to provide detailed information about their final submission as soon as possible.
</aside>
<aside> 🔔 [CLARIFICATION on June 18, 2024]
We want to clarify that each team has the freedom to decide whether or not to include figure indexes (e.g., "Figure 4." or "Fig 5.") in their output. For human evaluation, we will manually add or remove these indexes from your generated output to ensure that all captions are consistently presented either with or without indexes (depending on the final interface used for the evaluation) to human judges.
The captions in the test set on Eval.AI, which the leaderboard uses to calculate the percentages, do include indexes (e.g., "Figure 4."). Please note that we use the percentages calculated on the leaderboard only to determine if a submission qualifies for the Short/Long Caption Track, not to decide the winning teams.
</aside>
Join the 2nd Scientific Figure Captioning (SciCap) Challenge! We will supply each team with approximately 400,000 scientific figure images from various arXiv papers, including their respective captions and relevant paragraphs. Teams will then use these data to build computational models to generate captions for these images. Whether you are working alone or as a team, we welcome researchers, AI/NLP/CV practitioners, and anyone interested in computational models for generating useful text for visuals to participate and submit their results.
This year, the challenge will be hosted at IJCAI 2024 (August 3-9, Jeju Island, South Korea).
Check out the details of the challenge, including data, code, baselines, evaluation criteria, and important dates. We eagerly await your participation in the 2nd SciCap Challenge!
For questions about the challenge, email us at [email protected].
The organizers will host a 30-minute-long Zoom office hour to answer all kinds of questions. Please do not hesitate to join us!
The challenge will be structured into two tracks to accommodate the inherent differences and evaluation fairness between Short and Long captions. While longer captions tend to be viewed as more informative by readers [Huang et al. 2023], short captions are also crucial due to real-world space constraints in paper writing.