Another month is behind us, and once again the community came through with numerous advancements to the tooling and resources we need to read four scrolls this year. Congratulations to the following community contributors!
2 x $10,000 (Denarius)
Philip Allgaier (@spacegaier) has committed a laundry list of improvements to Volume Cartographer for the segmentation team:
mmap volume slice reading for performance and memory efficiency gains
Slice viewport rotation to ease neck strain; new shortcuts, settings, and defaults; dark mode, and more
Initial support for reading Zarr-format chunked volumes
These represent significant quality-of-life improvements to those using Volume Cartographer for heavy duty segmentation. Thank you, Philip!!
Chuck (@khartes_chuck) is also on a roll with contributions to another segmentation GUI, Khartes.
Cross-platform GPU support via OpenGL, enabling real-time rendering of segment surface volumes without pre-computation
3D segment implementation (rather than “2.5D”)
Chuck also shared a list of possible contributions others could make to Khartes
Khartes is steadily developing a very strong foundation for scroll data visualization and interaction, and is increasingly in use for segmentation (see below!). Thanks, Chuck!!
2 x $2,500 (Sestertius)
Sean Johnson (@Bruniss) has delivered yet another set of segments, this time from Scroll 2 - including the largest Scroll 2 segment to date! Could be of interest for a First Letters Prize… Other contributions include a preliminary segmentation idea and some data & labels masked to the region of the 2023 Grand Prize segments.
Johannes Rudolph (@jrudolph) developed vesuvius-gui, a Rust-based browser for scroll volume data. vesuvius-gui is capable of streaming chunked data real-time from a server, allowing one to explore the scroll without any up-front downloads or disk space required. Looking for an easy way to explore the datasets? This tool is quite simple to get up and running!
4 x $1,000 (Papyrus)
Oliver Daubney (@O_D) has been working on masking and compressing scroll data (discussion) for more accessibility.
James Darby (@james_darby) contributed a writeup and a repo investigating the tradeoffs of different volumetric network architectures for the segmentation of scrolls specifically.
Yao Hsiao (@Yao Hsiao) made a prototype browser-based scroll viewer, for viewing scroll data without downloading datasets or installing software.
Forrest McDonald (@VerditeLabs) added initial support for 3D segmentation in Hraun, allowing for arbitrary viewing and cutting angles:
Congratulations and thank you to all of this month’s prize winners!!
Weekly wrap-up
We’ve had a lot to share this week! Here’s the roundup:
There will be a breather before the next one. Not too long, though. These scrolls don’t read themselves!