📰 Medium · Machine Learning

By Day 5, I understood why WGAN uses a Critic instead of a traditional Discriminator. Continue reading on Medium »

Introduction: The Data Scarcity Bottleneck Continue reading on Medium »

How mobile vision models can map broken footpaths, illegal construction debris (Malba), and accessibility barriers in Indian metros. Continue reading on Medium

From self-driving cars to real-time surveillance, object detection is at the core of modern computer vision — and YOLOv8 is leading that… Continue reading on Me

A YOLOv8 + ByteTrack pipeline for real-time pedestrian counting, and the two bugs that almost made its numbers meaningless. Continue reading on Medium »

The CVPR 2026 Best Paper unifies depth, point clouds, camera pose, and 3D tracks behind one five-index interface. It gains speed by… Continue reading on Medium

You have a CPU you researched for three weeks. You have 32 gigs of RAM running at a speed you found in a forum post and now defend like… Continue reading on Med

Thousands of people. Hundreds of cameras. One question: Can technology identify you in a crowd? Continue reading on Medium »

Every CV project eventually reaches a moment where someone has to decide: is this good enough to ship? Not “is it perfect” — nothing is… Continue reading on Med

Learn how to build an object detection system that locates and classifies dice using CNN-based feature extraction, bounding box regression… Continue reading on

Building measurable spatial intelligence from monocular video with Depth Anything 3, SAM 3, calibrated cameras, and deterministic geometry Continue reading on M

Artificial Intelligence is no longer just about recognizing objects — it’s about understanding movement, behavior, and interaction. Continue reading on Medium »

As someone new to computer vision, I wanted to build a project that would expose me to the complete deep learning workflow — from… Continue reading on Medium »

In 2019, my final-year project partner, Ayesha Alhashemi, and I worked together to build a robot for our Bachelor of Software Engineering… Continue reading on M

Abstract. The forward pass turns pixels into a prediction; this final part runs the same path backwards — how a single number, the loss… Continue reading on Med

Why “Backbone,” “Embedding,” and “Prior” Feel Like a Foreign Language Even After Years in the Field Continue reading on Medium »

Why a good video remover spends most of its effort searching other frames for pixels it already has, and saves the imagination for last. Continue reading on Med

Learn what YOLOv8 is, why it was created, how it detects multiple objects in a single image, and how it differs from a traditional CNN —… Continue reading on Me

Artificial Intelligence is rapidly transforming the construction and interior design industries. From automated site monitoring to defect… Continue reading on M

This CVPR 2026 paper from an Oxford group tackles machine unlearning: removing a class or concept from an already-trained vision model… Continue reading on Medi

“Detecting an object is not always enough. Sometimes, the object needs to be clearer, cleaner, and easier to read before it can be… Continue reading on Medium »

You paid the flagship tax. Sixteen cores, a boost clock printed in bold on the box, a benchmark chart you screenshotted and sent to a… Continue reading on Mediu

In our previous deep dive, we explored Local Binary Pattern (LBP) — the lightweight champion of micro-texture analysis. Running relative… Continue reading on Me

You have a case with three intake fans you don’t need yet. A CPU cooler with a screen on it that displays your temperatures in a font you… Continue reading on M