<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Computer-Vision on CloudTales</title><link>https://cloudtales.gr/tags/computer-vision/</link><description>Recent content in Computer-Vision on CloudTales</description><generator>Hugo</generator><language>en-US</language><lastBuildDate>Fri, 20 Nov 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://cloudtales.gr/tags/computer-vision/index.xml" rel="self" type="application/rss+xml"/><item><title>Teaching a Model to See: Convolutional Networks in Five Small Programs</title><link>https://cloudtales.gr/convolutional-neural-networks-for-architects/</link><pubDate>Fri, 20 Nov 2026 00:00:00 +0000</pubDate><guid>https://cloudtales.gr/convolutional-neural-networks-for-architects/</guid><description>How image models actually look at a picture, shown with five small Python programs and their real output, and the shift test that drops a 90% model to 36%.</description></item></channel></rss>