<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
  <title>Ariful Kader — Blog</title>
  <link>https://thearifoffice.com/blog/</link>
  <description>Articles about AI engineering, computer vision, and building production systems</description>
  <language>en</language>
  <atom:link href="https://thearifoffice.com/rss.xml" rel="self" type="application/rss+xml"/>
  <item>
    <title>I Replaced a 730 MB Cartoon Model with a 78 MB One — and It Got Better</title>
    <link>https://thearifoffice.com/blog/cartoonify-journey/</link>
    <guid isPermaLink="true">https://thearifoffice.com/blog/cartoonify-journey/</guid>
    <description>How I went from a 730 MB pix2pixHD model that memorised its training data to a 78 MB U-Net that actually generalises — using FLUX as a teacher, a face refiner for small faces, and a lot of failed experiments.</description>
    <pubDate>Sat, 14 Nov 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>How I Solved YOLO Bounding Box False Positives with a Dual-Model Detection Strategy</title>
    <link>https://thearifoffice.com/blog/how-i-slove-yolo-bounding-box-false-positive/</link>
    <guid isPermaLink="true">https://thearifoffice.com/blog/how-i-slove-yolo-bounding-box-false-positive/</guid>
    <description>No single YOLO model gets bounding boxes right for every photo. Here's how running two models in parallel and merging their detections intelligently eliminated both false positives and missed detections in our production background removal service.</description>
    <pubDate>Tue, 10 Nov 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>Why I Build AI Systems That Ship, Not Just Demo</title>
    <link>https://thearifoffice.com/blog/why-i-build-ai-that-ships/</link>
    <guid isPermaLink="true">https://thearifoffice.com/blog/why-i-build-ai-that-ships/</guid>
    <description>Most AI projects die between prototype and production. Here's how I bridge that gap — from theme park photo engines to LCL freight tools.</description>
    <pubDate>Tue, 10 Nov 2026 00:00:00 GMT</pubDate>
  </item>
</channel>
</rss>
