Neurosymbolic Tag-Based Annotation for Interpretable Avatar Creation

Minghao Liu, Zeyu Cheng, Shen Sang, Jing Liu, James Davis

Keywords: Neurosymbolic AI, tag-based annotation, avatar creation

TL;DR: We use semantic tags as an intermediate representation between neural image understanding and symbolic avatar selection, achieving better annotation quality and model consistency than direct neural approaches.

Abstract: Avatar creation from human images presents challenges for direct neural approaches, which suffer from inconsistent predictions and poor interpretability due to the large parameter space with hundreds of ambiguous options. We propose a neurosymbolic tag-based annotation method that combines neural perceptual learning with symbolic semantic reasoning. Instead of directly predicting avatar parameters, our approach uses a neural network to predict semantic tags (hair length, curliness, direction) as an intermediate symbolic representation, then applies symbolic search algorithms to match optimal avatar assets. This neurosymbolic design produces higher annotator agreements (96.7\% vs 31.0\% for direct annotation), enables more consistent model predictions, and provides interpretable avatar selection with ranked alternatives. The tag-based system generalizes easily across rendering systems, requiring only new asset annotation while reusing human image tags. Experimental results demonstrate superior convergence, consistency, and visual quality compared to direct prediction methods, showing how neurosymbolic approaches can improve trustworthiness and interpretability in creative AI applications.

Track: Main Track

Paper Type: Long Paper

Resubmission: No

Submission Number: 12