2024 Grand Prize
The primary goal for 2024 is to read 90% of each of Scrolls 1-4, and we will issue the $100,000 2024 Grand Prize to the first team that is able to do this. In 2023, we got from 0% to 5% of one scroll. Reading 90% of each of our scrolls will lay the foundation to read all 300 scrolls.
More details on the exact judging criteria will be available in March, but you can start your work now. We will be looking for large, contiguous segmentations, possibly a single segmentation for the bulk of each scroll. The segmentations should pass geometric sanity checks, for example they cannot intersect themselves. We expect methods are going to need to operate in "true 3D" (as opposed to iterating through slices) and will need improved performance in compressed papyrus regions in order to meet this bar. We will also be looking for approaches that unwrap an entire segmented scroll into a contiguous 2D view, showing the original layout of the papyrus sheet. We anticipate that various tools in the pipeline will need updates to accommodate these larger segments.
2024 Progress Prize 1
Continuing our tradition of progress prizes, we’re awarding $25,000 in prizes (3 x $5,000, 5 x $2,000) for qualifying submissions by March 31st that get us closer to reading 90% of Scrolls 1-4. We will particularly favor progress on auto-segmentation, and tooling that allows people to compare ML outputs and transcriptions to the underlying data and improve labeling.
- Your submission must substantially increase the probability of reading 90% of our scrolls this year, as judged by the Review Team. We may award more or fewer prizes at our discretion, depending on the number of qualifying submissions.
- Your submission must be open source.
- We are favoring submissions that:
- Focus on auto-segmentation. In order to read an entire scroll and then multiple scrolls, we will need improved segmentation methods that operate in 3D, can handle compressed regions of papyrus, and result in large contiguous segmentations.
- Connect ink detection outputs with raw data. To improve ink detection, we believe it will be important to develop feedback loops between ink labeling, ink prediction, and papyrology. We need tools to do this, such as interfaces that allow one reading the ink predictions to also view the raw data and crackle patterns in place.
- Are released early. Tools released tomorrow have a higher chance of being used for reading the scrolls than those released a day before the deadline.
- Actually get used. We’ll look for signals from the community: questions, comments, bug reports, feature requests. Our Segmentation Team will publicly provide comments on tools they use.
- Submissions are closed on March 31st, 2024 at 11:59pm PT, after which the Review Team will select winners.
- Make your submission using this form.
More Progress Prizes
We have big plans for 2024, and you can expect to see more progress prizes throughout the year. We expect to invest heavily in the community this year. Let's make it another one to remember!