<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
	<channel>
		<title>BlazePose on 编程心语</title>
		<link>https://www.ithome.me/tags/blazepose/</link>
		<description>Recent content in BlazePose on 编程心语</description>
		<generator>Hugo</generator>
		<language>zh-CN</language>
		
		
		
		
			<lastBuildDate>Sat, 08 Aug 2026 08:00:00 +0800</lastBuildDate>
		
			<atom:link href="https://www.ithome.me/tags/blazepose/index.xml" rel="self" type="application/rss+xml" />
			<item>
				<title>端侧姿态估计实战：用 MediaPipe Pose Landmarker 给应用装上骨架追踪</title>
				<link>https://www.ithome.me/post/2026/08/08/on-device-ai-pose-estimation/</link>
				<pubDate>Sat, 08 Aug 2026 08:00:00 +0800</pubDate>
				<guid>https://www.ithome.me/post/2026/08/08/on-device-ai-pose-estimation/</guid>
				<description>&lt;h2 id=&#34;为什么要在端侧做姿态估计&#34;&gt;为什么要在端侧做姿态估计&lt;/h2&gt;&#xA;&lt;p&gt;健身 App 的深蹲计数、体感游戏的角色控制、运动康复的关节角度分析、儿童体态监测——这些场景的共同点是需要实时追踪人体骨架，而画面里全是敏感的身体信息。把姿态估计模型跑在设备端，视频不出手机，延迟低到能跟上 30fps 的视频流，还不用为每次推理付云服务费。这就是端侧姿态估计（On-Device Pose Estimation）的价值。&lt;/p&gt;</description>
			</item>
	</channel>
</rss>
