Embedded Files
Transform 2‑D biological images—captured by observers worldwide—into three‑dimensional scientific knowledge with your code.
Transform 2‑D biological images—captured by observers worldwide—into three‑dimensional scientific knowledge with your code.
This hackathon invites you to exploit an enormous archive of community‑science images and reconstruct unprecedented 3‑D views of living organisms. It goes far beyond conventional 3‑D model generation: at the frontier of AI × biology, we will integrate multi‑view data into genuine scientific insight. Merging open datasets with advanced AI is a hot topic for research teams everywhere. Bring your ideas and algorithms—let’s build the brand‑new Biological Encyclopedia 3.0 together!
Lead proposer of the selected idea & hackathon organizing member: Seine A. Shintani
NIBB‑CU Hackathon 2025: https://ideathonjp.nibb.ac.jp/
Award ceremony and official news
Award ceremony and official news
The award ceremony for the NIBB–Chubu University AI-Biology Hackathon 2025, “From Biological Images to 3D Models: Expanding Education and Research with AI,” was held on 22 May 2026 at Okazaki Conference Center.
The NIBB news article describes the hackathon as an implementation contest based on the award-winning idea proposed by Chubu University Associate Professor Seine A. Shintani, and introduces him as a member of the organizing team.
Official news:
Official news:
Tutorial available in the Discord community you can join after registering for the NIBB‑CU Hackathon 2025:
Tutorial available in the Discord community you can join after registering for the NIBB‑CU Hackathon 2025:
SPAR3D hands‑on session—turn a single image into a 3‑D model, even if you’re a complete beginner. ▶ Details
Kingfisher auto‑crop script: a program that generates two files from every photo▶ Details
Automatically detects the kingfisher in the picture and crops it—keeping the original resolution but including a small margin of background.
Produces an additional 512 × 512 px resized version of the cropped image.
Background Removal for Kingfisher Images▶ Details
Load crop_512.jpg (the 512 × 512 kingfisher crop) and run YOLOv8‑seg to segment the kingfisher at pixel level.
Automatically generate two files:
mask_kingfisher.png – a black‑and‑white mask of the kingfisher.
kingfisher_no_bg.png – a PNG with the background made transparent.
Page updated
Google Sites
Report abuse