Summary of Screenai: a Vision-language Model For Ui and Infographics Understanding, by Gilles Baechler et al.
ScreenAI: A Vision-Language Model for UI and Infographics Understanding
by Gilles Baechler, Srinivas Sunkara, Maria Wang, Fedir Zubach, Hassan Mansoor, Vincent Etter, Victor Cărbune, Jason Lin, Jindong Chen, Abhanshu Sharma
First submitted to arxiv on: 7 Feb 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- 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 ScreenAI is a vision-language model that specializes in understanding user interfaces (UIs) and infographics. The model improves upon the PaLI architecture by incorporating the flexible patching strategy of pix2struct and is trained on a unique mixture of datasets, including a novel screen annotation task. This task involves identifying the type and location of UI elements, which are then used to describe screens to Large Language Models and generate training datasets for question-answering (QA), UI navigation, and summarization tasks at scale. Ablation studies demonstrate the impact of these design choices on performance. At 5B parameters, ScreenAI achieves new state-of-the-art results on UI- and infographics-based tasks (Multi-page DocVQA, WebSRC, MoTIF, and Widget Captioning) and best-in-class performance on others (Chart QA, DocVQA, and InfographicVQA), compared to models of similar size. The model is released with three new datasets: one focused on the screen annotation task and two others focused on question answering. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary ScreenAI is a machine that can understand pictures and words. It’s good at understanding things like websites and diagrams. This helps computers talk to people better. To make it smart, we trained it on lots of examples and made it learn from mistakes. The new way we did this training helped the computer get even better! Now, ScreenAI is one of the best models for understanding pictures and words about websites and diagrams. |
Keywords
» Artificial intelligence » Language model » Question answering » Summarization
