Summary of Parsel: Parameterized Shape Editing with Language, by Aditya Ganeshan et al.
ParSEL: Parameterized Shape Editing with Language
by Aditya Ganeshan, Ryan Y. Huang, Xianghao Xu, R. Kenny Jones, Daniel Ritchie
First submitted to arxiv on: 30 May 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Graphics (cs.GR); Human-Computer Interaction (cs.HC); Symbolic Computation (cs.SC)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper presents a system called ParSEL that enables controlled editing of high-quality 3D assets from natural language inputs. The authors introduce a parameterized editing program that can be adjusted to explore shape variations with precise control. To generate these programs, they leverage large-language models (LLMs), but find that LLMs often fail to produce complete and semantics-preserving editing programs. To overcome this issue, the authors propose Analytical Edit Propagation (AEP), an algorithm that extends a seed edit with additional operations until a complete program is formed. AEP integrates computer algebra systems for geometric analysis to ensure analytical operations are compatible with various user edits. The authors experimentally demonstrate ParSEL’s effectiveness in enabling controllable 3D object editing through natural language requests, outperforming alternative system designs. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper creates a way to edit 3D objects using simple language. Right now, editing 3D assets is hard because we can’t easily describe what we want to change. The authors made a new system called ParSEL that lets us control how we edit these objects by giving it simple language instructions. They used special computer models to help them make the edits, but found that these models weren’t good at following the rules of editing 3D shapes. So they came up with a new way called Analytical Edit Propagation (AEP) that makes sure the edits are correct and follow the shape’s rules. The authors tested their system and showed it works better than other ways to edit 3D objects. |
Keywords
» Artificial intelligence » Semantics
