博文

目前显示的是 八月, 2026的博文

How to Recover Original Images from CDN Thumbnails: A Deep Dive into Image Processing Parameters

 Have you ever right-clicked an image on a website, only to find that the "original" you just saved is actually a tiny, pixelated thumbnail? Websites routinely serve compressed, resized versions of images to speed up page loading — but somewhere on the CDN, the full-resolution original still exists. The good news: in many cases, you can reconstruct or recover the original URL with just a few careful edits. Let's dig into how CDN image processing works and how to reverse it. ## The Problem: Thumbnails Everywhere When a website displays an image, the browser sends a request to the CDN. The CDN doesn't just serve the raw file — it applies a chain of image processing operations: resizing, cropping, compression, format conversion. For example, an Alibaba Cloud OSS image URL might look like this: https://img.example.com/photo/2024/cover.jpg?x-oss-process=image/resize,w_400/quality,q_75 This tells the OSS image processing service to resize the original to 400px wide and comp...

Running AI in the Browser: WebGPU-Accelerated MobileNet Feature Extraction for Real-Time Image Search

 Why Run Image Search in the Browser? When browsing image-rich websites, have you ever wanted to find similar images without uploading your pictures to a search engine? Traditional reverse image search relies on cloud APIs: upload images to a server, run deep learning models to extract features, then search a vector database. This approach has three major problems: 1. Privacy risk: your images must leave your device 2. Latency: network transfer plus server inference takes 1-3 seconds per query 3. Cost: API calls are metered, making large-scale use expensive Can we run the entire reverse image search pipeline in the browser? Yes — and with WebGPU, browser-side AI inference has reached practical performance levels. MobileNet: A Lightweight Network for Edge Devices To run deep learning in the browser, you first need a lightweight network architecture. MobileNet was designed for exactly this. Depthwise Separable Convolution MobileNet's core innovation is depthwise separable convolution...