Ancient Maya Writings as High-Dimensional Data: a Visualization Approach

Maya hieroglyphs

The ancient Maya civilization flourished from around 2000 BC to 1600 AD and left a great amount of cultural heritage materials behind, in the shape of stone monument inscriptions, folded codex pages, or personal ceramic items. All these materials have Mayan hieroglyphs, in short glyphs, written on them. The Maya writing system is visually complex and new glyphs are still being discovered. This brings the necessity of better digital preservation systems. Besides, the annotation of some glyphs is still open to discussion due to either visual differences or semantic analysis. Some glyphs are damaged, or have many variations due to artistic reasons, and to the evolving language.

Why producing spatial arrangements of glyphs?

Our visualizations are enabling (in some cases for the first time) scholars to discover meaningful patterns within the spatial arrangements into which large glyph datasets can configure themselves. It was not possible to quickly produce and reconfigure these arrangements for large glyph datasets prior to the advent of Digital Palaeography methods, although it is known that earlier Catalog authors such as Thompson or Zimmermann relied on graphic cards to study similar patterns and spatial distributions.

Studying spatial arrangements of glyphs could ultimately reveal crucial links between semantic and visual similarities which could help to overcome the limitations of existing glyph-catalogues and reveal basic principles of ancient Maya sign categorization which have escaped scholars in the past. An approach emphasizing the semantic similarity with the support of the visually similarity probably turn out to be more productive. In this way semantic "regions" can be build, where notions such as "colors", "cardinal directions" and specific toponyms from earthly, heavenly or underworldy realms begin to overlap.These are the kind of associations that such a visualization system can facilitate to study in greater detail.


In this work, we used a bag-of-words (BoW) representation for glyph based on patch-based 2-ring HOOSC descriptors computed at sample glyph positions (we tested 4 different spatial context sc for this local descriptors; sc = 1/1 corresponds to 128x128 patch for computing the feature). As for the vizualization, we relied on the t-distributed Stochastic Neighborhood Embedding (t-SNE) methodology applied to these BoW representations.

Please click on the links to see visualization of glyph samples.

Explanation of visual clusters

In this video, we explain our methodology, the glyph datasets we used, and our observations/interpretations of the visualizations.

Ancient Maya Writings as High-Dimensional Data: a Visualization Approach,
Gulcan Can, Jean-Marc Odobez, Carlos Pallan Gayol and Daniel Gatica-Perez, in: Digital Humanities, 2016.

author = {Can, Gulcan and Odobez, Jean-Marc and Pallan Gayol, Carlos and Gatica-Perez, Daniel},
projects = {Idiap, MAAYA},
title = {Ancient Maya Writings as High-Dimensional Data: a Visualization Approach},
booktitle = {Digital Humanities},
year = {2016}

Go Back to Project Website