ProtoSnap: Prototype Alignment for Cuneiform Signs

Given a target image of a cuneiform sign, and a corresponding prototype, we align the skeleton with the target image ('snapping' the prototype into place)

ProtoSnap: Prototype Alignment for Cuneiform Signs ProtoSnap: Prototype Alignment for Cuneiform Signs Rachel Mikulinsky* 1 Morris Alper* 1 Shai Gordin 2 Enrique Jimenez 3 Yoram Cohen 1 Hadar Averbuch-Elor 1,4 1 Tel Aviv University 2 Ariel University 3 LMU 4 Cornell University * Equal Contribution ICLR 2025 arXiv Code TL;DR: Given a target image of a cuneiform sign, and a corresponding prototype, we align the skeleton with the target image ("snapping" the prototype into place). Abstract --> The cuneiform writing system served as the medium for transmitting knowledge in the ancient Near East for

saved by

• David Oks

related reading

• Sequence Modeling with CTCdistill.pub
• Structure and Interpretation of Deep Networkssidn.baulab.info
• [2205.13147] Matryoshka Representation Learningarxiv.org
• [2603.05923] Learning Next Action Predictors from Human-Computer Interactionarxiv.org
• [1912.13318] LayoutLM: Pre-training of Text and Layout for Document Image Understandingarxiv.org
• Position: It's Time to Optimize for Self-Consistencytime-for-consistency.github.io
• Kimi Linear: An Expressive, Efficient Attention Architecturearxiv.org
• [2108.12409] Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolationarxiv.org
• [1711.08189] An Analysis of Scale Invariance in Object Detection - SNIParxiv.org
• Superquadrics Revisited: Learning 3D Shape Parsing beyond Cuboidsarxiv.org
• [2602.15799] The Geometry of Alignment Collapse: When Fine-Tuning Breaks Safetyarxiv.org
• Telling Left from Right: Identifying Geometry-Aware Semantic Correspondencetelling-left-from-right.github.io