📰 Medium · Python

Open any self-driving demo and somewhere underneath there is a neural network the size of a small country, trained on more video than you… Continue reading on M

In my recent journey building an automated anomaly detection system for security footage, I hit a wall. I had a model that performed well… Continue reading on M

Default overlays look like debug output Continue reading on Medium »

When I first started learning about Artificial Intelligence, I thought it was all about robots, chatbots, and complex programming. During… Continue reading on M

A step-by-step breakdown of building a pose analysis pipeline with just a phone camera and no hardware of my own — what broke, what the… Continue reading on Med

Last week I typed “a black backpack” into a search box and watched a backpack light up yellow inside a 3D reconstruction of my apartment… Continue reading on Me

This playbook serves as a comprehensive reference guide for training, fine-tuning, and deploying a hybrid Deep Learning model (MobileNetV2… Continue reading on

Sebuah pendekatan praktis pengolahan citra digital untuk otomatisasi segmentasi gambar produk dan profil tanpa perangkat lunak… Continue reading on Medium »

How our team built a real-time Computer Vision system using YOLO, OpenCV, and DeepFace to assist professional certification exams and why… Continue reading on M

A step-by-step guide to deploying your first Computer Vision AI project using OpenCV and Streamlit. Continue reading on Medium »

“Clifford Vortex Filaments” is a generative media art piece that dives deep into chaos theory and non-linear dynamics, visualizing a… Continue reading on Medium

Imagine capturing a near-perfect cinematic shot, only to notice an ugly corporate logo, a distracting background object, or an accidental… Continue reading on M

Build a live skeleton overlay that detects up to two hands, colors each finger differently, and runs at smooth frame rates — in under 140… Continue reading on M

In the late autumn of 1889, Vincent van Gogh stood before his easel in Saint-Rémy-de-Provence, applying thick, deliberate strokes of… Continue reading on Medium

How I built a full-stack automated vehicle compliance system to track plates, process video feeds, and issue automated warnings. Continue reading on Medium »

A hands-on walkthrough of building, training, and evaluating a Convolutional Neural Network (CNN) on the Fashion-MNIST dataset using… Continue reading on Medium

And it spots patterns your nutritionist would catch — for $5/month Continue reading on Medium »

Building BrandSpotter: a three-stage brand recognition pipeline on LogoDet-3K, and why the hardest part wasn’t detection or classification. Continue reading on

Di artikel kali ini, saya mau sharing sebuah side project gabut tapi berfaedah yang baru-baru ini saya selesaikan. Sebagai mahasiswa yang… Continue reading on M

I built an interactive CNN explainer that lets you visually see how a Convolutional Neural Network processes images — layer by layer. Continue reading on Medium

Food image understanding is usually treated as a classification problem. Given an image, the model predicts one label such as pizza… Continue reading on Medium

Crowd counting is a very sophisticated problem in computer vision. Standard object detectors easily fail when people stand too close… Continue reading on Medium

In computer vision, the Scale-Invariant Feature Transform (SIFT) algorithm remains a classic foundational standard for keypoint detection… Continue reading on M

From filters to feature maps: building networks that actually see. Continue reading on Medium »