<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>屏幕注释 on AI 早报</title><link>https://ai-news.example.com/tags/%E5%B1%8F%E5%B9%95%E6%B3%A8%E9%87%8A/</link><description>Recent content in 屏幕注释 on AI 早报</description><generator>Hugo</generator><language>zh-CN</language><lastBuildDate>Mon, 06 Jul 2026 03:29:01 +0800</lastBuildDate><atom:link href="https://ai-news.example.com/tags/%E5%B1%8F%E5%B9%95%E6%B3%A8%E9%87%8A/feed.xml" rel="self" type="application/rss+xml"/><item><title>[EN]Google提出ScreenAI：统一理解UI与信息图的视觉语言模型</title><link>https://ai-news.example.com/articles/engooglescreenaiui/</link><pubDate>Mon, 06 Jul 2026 03:29:01 +0800</pubDate><guid>https://ai-news.example.com/articles/engooglescreenaiui/</guid><description>Google研究人员推出ScreenAI，一种基于PaLI架构与pix2struct灵活分片策略的视觉语言模型，仅5B参数即可在UI与信息图理解任务上达到业界领先水平。模型通过自监督预训练与人工标注微调，结合大规模屏幕标注数据，能完成UI元素识别、导航、问答等任务，并开源了三个新数据集。</description></item></channel></rss>