Daejeon, July 27–31
The use of 3D technologies in the humanities has increasingly become a subject of scholarly discussion in recent years (Münster 2022), alongside significant advances in data creation, management, and exhibition. Many universities and GLAM institutions have begun developing 3D collections, yet robust frameworks for annotating, semantically enriching, and reusing these resources remain underdeveloped.
In response to this situation, this research proposes a comprehensive framework for authoring, semantically enriching, and visualizing 3D annotations within the IIIF and Linked Data ecosystem. The work began with the development of an annotation authoring environment that enables users to define spatial regions on 3D models—represented as points or polygonal areas—and associate them with external knowledge sources such as Wikidata, bibliographic metadata, images, and video materials available online.
Within this framework, annotations are treated as addressable conceptual entities identified by URIs and described using RDF-compatible data structures, allowing them to participate in Linked Data–oriented knowledge infrastructures. To support visualization and reuse, a complementary IIIF 3D Viewer was subsequently developed to render 3D models and their enriched annotations encoded in extended IIIF Manifests. Together, the authoring environment and viewer establish an end-to-end workflow for creating, publishing, and reusing semantically enriched 3D cultural heritage data.
In the cultural heritage domain, platforms such as Sketchfab are widely adopted and provide basic annotation functionalities; however, they do not interoperate with international standards and lack standardized, shareable annotation structures suitable for scholarly reuse.
More advanced approaches incorporating storytelling mechanisms—such as the PURE3D project—have also attracted attention (Papadopoulos & Schreibman 2025). Nevertheless, despite these developments, the implementation of 3D annotation within widely adopted digital humanities frameworks, including IIIF, TEI, and RDF, remains extremely limited.
Parallel efforts to integrate 3D cultural heritage data with Linked Data infrastructures can be found in platforms such as Kompakkt and Semantic Kompakkt, which leverage Wikibase-based knowledge management systems [1]. These initiatives demonstrate the value of embedding 3D objects within knowledge graphs, but they do not provide interoperability with IIIF, nor do they generate IIIF Manifests or annotation structures consumable by IIIF-compliant viewers. Moreover, 3D models are generally treated as monolithic entities, limiting fine-grained semantic modeling of discrete spatial regions.
The Virtual Interiors project represents one of the most comprehensive examples of a fully developed 3D platform integrating Linked Data (Huurdeman & Piccoli 2021). In this project, each 3D object is semantically linked via RDF to related documents, images, and external data sources, enabling interactive access through the user interface. Despite its significance, the number of annotated models remains limited, the annotation process is largely unidirectional and controlled by data providers, and the sharing of 3D data and annotation information does not necessarily conform to internationally recognized standards.
The system presented here builds upon established standards and technologies, including the IIIF Presentation API, the Web Annotation Data Model extended to support 3D spatial targets, RDF-compatible representations for semantic modeling, and WebGL/Three.js for browser-based visualization. By integrating these components, the system provides a practical and interoperable approach to 3D annotation within the IIIF and Linked Data ecosystem.
IIIF (The International Image Interoperability Framework) has become a foundational standard for the interoperable dissemination, preservation, and scholarly use of cultural heritage materials. While IIIF is now widely adopted for images, audiovisual media, and textual annotations, its application to three-dimensional cultural objects remains at an early stage. Although experimental methods for representing 3D resources in IIIF Manifests have been proposed (IIIF 3D Technical Specification Group 2025), standardized approaches for encoding spatial annotations, semantically enriching them with external knowledge, and integrating them into existing IIIF workflows are still underdeveloped. As a result, 3D cultural heritage objects often remain isolated from the interoperable infrastructures that already support images and texts, limiting their scholarly reuse and comparative analysis.
At the core of this research is a web-based environment that enables users to author semantically enriched 3D annotations in a manner compatible with IIIF [2]. Users load a IIIF Manifest describing a 3D model and define annotation targets by selecting points or drawing polygonal regions directly on the model’s surface. Each annotation can be associated with descriptive text, images, audiovisual materials, bibliographic references, and external identifiers such as Wikidata items as shown in Figure 1.
A distinctive feature of the system is its ability to automatically retrieve and integrate external semantic information from Linked Data sources. When an annotation references a Wikidata entity that contains geographic information—such as coordinates or place identifiers—the system automatically extracts this information and attaches it to the annotation as structured metadata. This enriched metadata is embedded in the extended IIIF Manifest generated by the system using RDF-compatible representations [3].
The annotation data model extends the IIIF Georeference Extension and Linked Places Format (LPF) to 3D spatial contexts by introducing coordinates that map geographic locations to positions on 3D model surfaces (Grossner & Mostern 2021; IIIF Maps Technical Specification Group 2023), enabling synchronized visualization across map and 3D viewer interfaces as shown in Figure 2. The accompanying IIIF 3D Viewer interprets these enriched Manifests and enables annotations to be visualized across multiple representational modalities.
This functionality illustrates how Linked Data enrichment allows annotations anchored in 3D object space to be connected to real-world geographic contexts, supporting multiscale analysis that links object-level features with broader spatial and historical frameworks. The annotations authored in the system can be exported in two interoperable forms:
This research presents a unified and interoperable framework for authoring, semantically enriching, visualizing, and disseminating 3D annotations within IIIF and Linked Data–oriented infrastructures. By enabling automated semantic enrichment and supporting annotation-level discovery across collections, the system shifts access to 3D cultural heritage data from an object-centered model toward a more granular, knowledge-driven approach. These contributions provide both a technical implementation and a conceptual foundation for future IIIF 3D standardization efforts and open new possibilities for comparative, contextual, and cross-collection analysis of 3D cultural heritage materials.
The system has been applied to annotated 3D Latin inscriptions, demonstrating its ability to support semantic queries across objects based on iconographic and conceptual features. Preliminary use confirms that linking annotations to external knowledge based such as Wikidata enables cross-objects semantic discovery and onteroperability between IIIF-based visualization and RDF-based semantic querying [4].
[1] Semantic Kompakkt: https://kompakkt.de/about
[2] https://3-d-annotation-viewer.vercel.app/editor/3d. Currently, only the glTF/GLB format is supported for 3D models referenced by uploaded IIIF Manifests.
[3] An extended IIIF Manifest for the given IIIF 3D manifest (https://3d-iiif-viewer.vercel.app/manifests/sample-manifest.json) is provided below: https://3-d-annotation-viewer.vercel.app/api/3/aHR0cHM6Ly8zZC1paWlmLXZpZXdlci52ZXJjZWwuYXBwL21hbmlmZXN0cy9zYW1wbGUtbWFuaWZlc3QuanNvbg/manifest
[4] If “animal” is entered in the “Wikidata Search” field of the search system below, 3D objects whose annotations are linked to the corresponding Wikidata entity for “animal” are retrieved: https://3-d-annotation-viewer.vercel.app/search