DigitCode: Symbolic Tokenization of Hand Motion by Anatomical Units

2026-08-04

AI Summary

Current representations of hand motion typically rely on continuous parameters, lacking a structured symbolic system that enables efficient indexing, editing, and adherence to anatomical constraints. To address this, this work proposes DigitCode—a hierarchical symbolic encoding framework grounded in the anatomical hierarchy of the hand (bone → finger → whole hand)—which represents hand movements as discrete, interpretable symbol sequences. Integrating the Hand Labanotation symbol system, the approach introduces both training-agnostic and learning-based strong quantizers and establishes the HandTok evaluation platform. The method reduces quantization error by 75% while preserving symbolic interpretability, and its single-finger symbols are directly applicable to hand pose correction and robotic retargeting, demonstrating the critical role of anatomical units in downstream tasks.

Abstract

Hand motion carries the finest-grained information in human activity, yet the representations behind hand generation, understanding, and robot learning are overwhelmingly continuous--joint angles or MANO parameters. These are accurate but unstructured: a finger cannot be indexed or edited as a symbol, and nothing marks a pose as anatomically valid. Discrete symbolic representations supply exactly this structure, and Hand Labanotation (HL) has shown they are feasible for the hand, writing motion as a T x 40 grid of one fixed direction symbol per bone. Building on this grid, we ask the question underneath it: the anatomical unit a symbol should span--bone, finger, or whole hand. DigitCode answers it by adapting, grouping, and layering HL's alphabet along the hand's unit hierarchy within one code, cutting the symbolic representation's quantization error by three quarters. The lever is the unit, not the quantizer family: at a fixed unit, training-free and learned strong quantizers are interchangeable on reconstruction, while moving down the anatomical hierarchy is what shifts accuracy. The hierarchy also tracks what downstream tasks need. Because a finger is a genuine, enumerable unit, one per-finger token doubles as a training-free, editable handle for jobs a continuous representation cannot address--repairing malformed generated hands, and retargeting them onto robots. We release HandTok, a reproducible testbed, so hand tokenizers can be compared unit-for-unit. Project page: https://digitcode-demo.github.io.

Problem

• hand motion representation
• symbolic tokenization
• anatomical units
• discrete representation
• hand generation

Innovation

• symbolic tokenization
• anatomical hierarchy
• hand motion representation
• discrete hand coding
• DigitCode