Generative Shape Editing via SDF Handle Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional graphics editing systems face challenges in accurately representing and editing shapes due to their inability to capture shape structure effectively, leading to inaccuracies and inefficiencies in user interactions and system resource utilization.
Innovation Solution
The system represents training shapes as sets of visual elements called 'handles' and converts them into signed distance field (SDF) representations, training a handle processor model to generate new shapes that reflect salient visual features, enabling accurate shape characterization and editing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional graphics editing systems use deep generative models to learn distributions of raw shape representations (occupancy grids, point clouds, meshes), then shape generation capability is improved, but shape structure representation accuracy deteriorates and editability is lost
Solution Approach 1:
The patent introduces SDF representations as an intermediary between raw shape data and editable shape representations. The SDF serves as a mediator that captures accurate shape structure while enabling subsequent handle-based editing operations, resolving the contradiction between generation capability and structure accuracy.
Solution Approach 2:
The patent segments the shape representation into two parts: SDF representation for accurate structure capture and handle representations for editable geometric control. This segmentation allows each component to optimize for its specific function without compromising the other.
2Ease of operation
If shapes are represented using compact representations (cages, skeletons, primitives, curves), then ease of editing is improved, but shape structure accuracy deteriorates
Solution Approach 1:
The SDF acts as an intermediary layer that maintains accurate shape structure information while allowing handle-based editing operations. The SDF ensures that edits to handles preserve structural accuracy, resolving the contradiction between ease of editing and structure accuracy.
3Ease of operation
If inaccurate shapes are generated by conventional systems, then user interaction complexity increases, but system resource consumption increases
Solution Approach 1:
The system performs preliminary action by generating accurate SDF representations and corresponding handles before user editing begins. This preliminary accuracy reduces the need for multiple iterative corrections, thereby reducing overall system resource consumption while maintaining ease of operation.
Data Source
AI summary
Generative shape creation and editing is leveraged in a digital medium environment. An object editor system represents a set of training shapes as sets of visual elements known as “handles,” and converts sets of handles into signed distance field (SDF) representations. A handle processor model is then trained using the SDF representations to enable the handle processor model to generate new shapes that reflect salient visual features of the training shapes. The trained handle processor model, for instance, generates new sets of handles based on salient visual features learned from the training handle set. Thus, utilizing the described techniques, accurate characterizations of a set of shapes can be learned and used to generate new shapes. Further, generated shapes can be edited and transformed in different ways.