Skip to main content

Series

Computer Vision Engineering

Computer Vision Engineering

Welcome to the Computer Vision Engineering series by Biswas Labs.

Computer Vision is one of the most exciting areas of Artificial Intelligence, enabling machines to interpret, analyze, and understand visual information.

This series documents my journey of exploring computer vision, from understanding fundamental concepts and deep learning architectures to implementing practical systems that solve real-world problems.

Rather than focusing only on theoretical explanations, I aim to connect mathematical intuition, architectural understanding, implementation, experimentation, and deployment.

What You'll Explore

  • Computer Vision Fundamentals: Image processing, image representations, feature extraction, and visual understanding.
  • Deep Learning Architectures: CNNs, encoder-decoder architectures, SegNet, U-Net, and other modern vision architectures.
  • Object Detection: YOLO architectures, bounding-box regression, object localization, and tracking.
  • Image Segmentation: Semantic segmentation, instance segmentation, and practical implementations.
  • Model Optimization: Quantization, computational complexity, inference latency, and memory optimization.
  • Real-World Applications: Agricultural AI, intelligent detection systems, automation, and edge deployment.

Learning Philosophy

Understanding an architecture is not the same as being able to implement it.

Throughout this series, I will focus on understanding how models work internally, why specific architectural decisions are made, how to implement them, and how to evaluate their performance.

The goal is to move beyond simply using existing libraries and develop the ability to understand, build, experiment with, and deploy computer vision systems.

Let's explore how machines learn to see, one project at a time.

— Biswas Pokhrel | Biswas Labs

Posts in this series