Medium

Md Faruk Alam

Md Faruk Alam

·

Sep 16

Object Detection and Tracking in a Simulated Warehouse with NVIDIA Isaac Sim, RF-DETR, and ROS 2

A step by step guide: a custom-trained detection transformer watching a simulated warehouse camera over ROS 2 Jazzy, with live tracking in…

Object Detection and Tracking in a Simulated Warehouse with NVIDIA Isaac Sim, RF-DETR, and ROS 2

In

cvRealtime

by

Md Faruk Alam

·

Aug 28

Training a Simulated SO-ARM101 Robot Arm to Lift a Cube with NVIDIA Isaac Lab

A complete walkthrough of reinforcement learning in Isaac Sim on a single 16GB GPU, including every error hit along the way and how to fix…

Training a Simulated SO-ARM101 Robot Arm to Lift a Cube with NVIDIA Isaac Lab

In

cvRealtime

by

Md Faruk Alam

·

Jul 24

RF-DETR CUDA (Part 2): Bilinear Sampling With Texture Objects

Part 1 of this series built an image preprocessing kernel where every output pixel maps to a predictable source location on a regular grid…

RF-DETR CUDA (Part 2): Bilinear Sampling With Texture Objects

In

cvRealtime

by

Md Faruk Alam

·

Jul 13

RF-DETR CUDA (Part 1): Building a Fused Image Preprocessing Kernel from Scratch

This is Part 1 of a series where I implement the compute-critical parts of RF-DETR — a state-of-the-art real-time object detector — in CUDA…

RF-DETR CUDA (Part 1): Building a Fused Image Preprocessing Kernel from Scratch

In

cvRealtime

by

Md Faruk Alam

·

Apr 8

RF-DETR’s New PyTorch Lightning Training Stack

RF-DETR rewrote its entire training stack on PyTorch Lightning. If you’ve been training object detection models with custom loops, this is…

RF-DETR’s New PyTorch Lightning Training Stack

In

cvRealtime

by

Md Faruk Alam

·

Jan 8

Achieving 374 FPS with YOLOv8 Segmentation on NVIDIA RTX 5070 Ti GPU

In this comprehensive guide, I’ll walk you through optimizing YOLOv8 segmentation models with NVIDIA TensorRT, achieving impressive

Achieving 374 FPS with YOLOv8 Segmentation on NVIDIA RTX 5070 Ti GPU

In

cvRealtime

by

Md Faruk Alam

·

Jan 1

Choosing the Right Edge AI Hardware for Your 2026 Computer Vision Application

The edge computing landscape has transformed dramatically over the past year. With platforms like NVIDIA Jetson Thor delivering…

Choosing the Right Edge AI Hardware for Your 2026 Computer Vision Application

In

cvRealtime

by

Md Faruk Alam

·

Dec 17, 2025

Focus on Your Algorithm: A First Look at NVIDIA CUDA Tile for Computer Vision

If you have spent any time optimizing computer vision pipelines, you know the specific kind of headache that comes with writing custom CUDA…

Focus on Your Algorithm: A First Look at NVIDIA CUDA Tile for Computer Vision

In

cvRealtime

by

Md Faruk Alam

·

Nov 29, 2025

✍️ Become a Writer for cvRealtime

Welcome to cvRealtime, a community-driven publication dedicated to helping developers learn and build real-time computer vision…

✍️ Become a Writer for cvRealtime

In

cvRealtime

by

Md Faruk Alam

·

Nov 4, 2025

Object Detection with Multimodal Large Vision-Language Models: An In-depth Review

The world of computer vision is experiencing a revolutionary transformation. For years, object detection systems have relied on…

Object Detection with Multimodal Large Vision-Language Models: An In-depth Review

Md Faruk Alam

Md Faruk Alam

Computer Vision Engineer | Machine Learning | Deep Learning | Vision Language Models Subscribe to my Newsletter: https://farukalamai.substack.com/

Following

Help

Status

About

Careers

Press

Blog

Store

Privacy

Rules

Terms

Text to speech