Avatar Concept Slider: Controllable Editing of Concepts in 3D Human Avatars

2024-08-26

AI Summary

To address insufficient editing precision in 3D human avatar generation caused by natural language ambiguity, this paper proposes a semantics-aware controllable editing framework. Our method constructs interpretable semantic concept axes (e.g., “young ↔ old”) via Latent Dirichlet Allocation (LDA), constrains the identity-preserving feature space using Principal Component Analysis (PCA), and jointly optimizes geometry and appearance through 3D Gaussian splatting. We introduce two novel loss terms—concept-sliding loss and attribute-preservation loss—for collaborative optimization, and design a concept-sensitivity-driven Gaussian point selection and update strategy to enhance editing accuracy and efficiency while maintaining identity consistency and visual fidelity. Experiments demonstrate that our approach enables fine-grained, interpolatable, identity-invariant, high-quality editing across multiple semantic dimensions, significantly outperforming existing text-driven and latent-space editing methods.

Technology Category

• Computer Vision: 3D Computer Vision
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Natural Language Processing: Generation

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Text-based editing of 3D human avatars to precisely match user requirements is challenging due to the inherent ambiguity and limited expressiveness of natural language. To overcome this, we propose the Avatar Concept Slider (ACS), a 3D avatar editing method that allows precise editing of semantic concepts in human avatars towards a specified intermediate point between two extremes of concepts, akin to moving a knob along a slider track. To achieve this, our ACS has three designs: Firstly, a Concept Sliding Loss based on linear discriminant analysis to pinpoint the concept-specific axes for precise editing. Secondly, an Attribute Preserving Loss based on principal component analysis for improved preservation of avatar identity during editing. We further propose a 3D Gaussian Splatting primitive selection mechanism based on concept-sensitivity, which updates only the primitives that are the most sensitive to our target concept, to improve efficiency. Results demonstrate that our ACS enables controllable 3D avatar editing, without compromising the avatar quality or its identifying attributes.

Problem

• Enables precise text-based editing of 3D human avatars.
• Introduces a slider mechanism for semantic concept adjustment.
• Improves efficiency by updating only concept-sensitive primitives.

Innovation

• Concept Sliding Loss for precise editing
• Attribute Preserving Loss maintains avatar identity
• 3D Gaussian Splatting improves editing efficiency