📄 中文摘要
在边缘人工智能的时代，隐私和性能不再是权衡的选择。用户对上传敏感健康数据（如皮肤影像）到中央服务器越来越谨慎。通过在浏览器中实现计算机视觉，能够有效解决这一问题。利用 TensorFlow.js、MediaPipe 和 WebGPU，可以构建一个以隐私为首的人工智能应用，直接在客户端设备上进行实时皮肤病变分割和特征提取。所有数据均不离开浏览器，得益于硬件加速，推理速度极快。该教程将探讨如何结合 MediaPipe 的结构性优势与分类能力，构建高效的皮肤健康监测工具。
边缘计算下的皮肤健康：使用 MediaPipe 和 TensorFlow.js 进行实时病变筛查 🩺✨
📄 English Summary
Skin Health at the Edge: Real-time Lesion Screening with MediaPipe and TensorFlow.js 🩺✨
In the era of Edge AI, privacy and performance are no longer a trade-off. Users are increasingly cautious about uploading sensitive health data, such as skin imaging, to a central server. Implementing computer vision in the browser addresses this concern effectively. By leveraging TensorFlow.js, MediaPipe, and WebGPU, a privacy-first AI application can be built that performs real-time skin lesion segmentation and feature extraction directly on the client’s device. No data leaves the browser, and inference is lightning-fast due to hardware acceleration. This tutorial explores how to combine the structural power of MediaPipe with classification capabilities to create an efficient skin health monitoring tool.