actFig

actFig is a standardized data model and file format designed to represent articulated figures and their actions for use in computer graphics, robotics, and biomechanics. The format aims to enable interoperable storage and exchange of skeleton definitions, joint parameters, pose states, and temporal motion data across diverse software tools and pipelines.

Origin and evolution: actFig was conceived by the ActFig Consortium, a collaboration of researchers and industry developers, with the goal of reducing fragmentation in figure data used for animation, simulation, and analysis. The initial specification was released publicly in 2021, followed by incremental revisions and reference implementations in multiple programming languages. The format is open and published under a permissive license.

Structure and content: An actFig file consists of metadata, a figure description, pose data, and motion tracks. The figure description encodes a skeleton or hierarchy of articulated parts; joints have defined axes, limits, and base transforms. Poses describe static configurations, while motion tracks store time-stamped samples such as joint angles or global poses. ActFig supports multiple figures per file, versioning, units, coordinate conventions, and provenance information.

Encoding and interoperability: actFig is offered in a human-readable JSON-based schema and an accompanying compact binary form for efficient parsing. It provides converters and adapters to other popular formats, including BVH, URDF, FBX, and GLTF, enabling import and export between pipelines. Adoption is strongest in research, robotics simulation, and game animation workflows.

Limitations and reception: While growing in popularity, actFig remains less entrenched than legacy formats in certain industries, and ongoing work addresses non-rigid deformations, complex constraints, and scene metadata alignment.