AIKON enables targeted processing of document sets using computer vision algorithms. Users select documents, configure treatments, and review results through dedicated interfaces.
A document set is a personal collection of witnesses, works, and series. Sets group materials for processing according to your research logic.
Treatments represent tasks applied to document sets. AIKON handles three types of processes:
Track treatment status in the treatment list interface. Processing time depends on document count and algorithm complexity.
After region extraction, navigate to a witness to view detected regions. Two views are available:
AIKON provides tools to correct automated detections:
The similarity interface displays ranked matches for each region. Researchers can classify correspondences and build validated networks.
Annotated matches enable propagation: if A matches B and B matches C, then A and C form a propagated similarity. This helps reveal annotation inconsistencies and discover new connections.
Use the toolbar to filter results by score, witness, and similarity level. The detailed view (eye button) allows overlaying images for precise comparison.