Summary of Sketch2code: Evaluating Vision-language Models For Interactive Web Design Prototyping, by Ryan Li et al.
Sketch2Code: Evaluating Vision-Language Models for Interactive Web Design Prototyping
by Ryan Li, Yanzhe Zhang, Diyi Yang
First submitted to arxiv on: 21 Oct 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
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 The proposed Sketch2Code benchmark evaluates the ability of Vision Language Models (VLMs) to convert simple sketches into webpage prototypes. The model is tested on its ability to refine its generations through interactive communication with a simulated user, either receiving feedback or asking clarification questions. Ten commercial and open-source VLMs are analyzed, showing that even the most capable models struggle to accurately interpret sketches and formulate effective questions for improvement. A user study with UI/UX experts reveals a preference for proactive question-asking over passive feedback reception, highlighting the need for more effective paradigms. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Sketch2Code is a new way to test how well computers can turn simple drawings into website designs. It lets VLMs (computer models) try to make webpage prototypes from sketches and then get better at it by asking questions or getting feedback. Ten different computer models were tested, but even the best ones struggled to do this task. People who design websites like this prefer when the computer asks them questions instead of just giving them information. |
