Categories
Enroll now

Course Details

Computer Vision from Fundamentals to Practical

Data Science
course-meta
Created by

Last Update

September 13, 2023

Created On

July 01, 2023

Description

Computer Vision is an AI field that empowers computers to extract valuable insights from digital images, videos, and visual inputs, enabling them to make informed decisions and recommendations. It provides machines with the ability to observe, understand, and interpret visual data. With applications spanning various industries, including energy, utilities, manufacturing, and automotive, the field of computer vision is experiencing ongoing growth and development.

Overview

" Computer Vision from Fundamentals to Practical " is an intensive program that deepens participants' knowledge of computer vision, from fundamentals to practical applications. This course covers key topics like CNNs, object detection, image segmentation, and hyperparameter tuning, providing hands-on experience with industry-standard tools like TensorFlow and PyTorch. By course end, participants will have the expertise to address complex computer vision challenges and contribute to innovative solutions in fields like healthcare and autonomous vehicles.

Features

  • Comprehensive Content
  • Hands-on Practical Experience
  • Variety of Architectures CNN architectures, including LeNet-5, AlexNet, VGGNet, Inception, ResNet, and YOLO
  • Hands-on Practicals
  • Challenges
  • Downloadable resources
  • Quizzes
  • GPU Cloud Providers
  • GANs Introduction
  • Practical Projects
  • Completion Certificate

What you'll learn

  • Advanced Techniques in Computer Vision
  • Deep Learning for Computer Vision
  • Convolutional Neural Networks (CNN) in Computer Vision
  • Object Detection and Segmentation in Computer Vision
  • Image Formation and Processing in Computer Vision
  • Motion Estimation and Tracking in Computer Vision
  • Advanced Applications of Computer Vision Architectures
  • Implementing Computer Vision for Advanced Projects.

Prerequisites

Curriculum

  • 25 modules

Course Overview and Objectives

Course Outcome and Expected Learning

Installing Anaconda, Pycharm, and Postman

Setting up and Managing Conda Environments

Introduction to Pycharm IDE

Integrating Pycharm with Conda

Integrating Pycharm with Virtual Environments (venv)

Integrating Pycharm with Pipenv

The Importance of CNN: A Comprehensive Overview

CNN Components: Kernels, Channels, Feature Maps, Stride, and Padding

Receptive Fields and Image Output Dimensionality Calculations

Exploring the MNIST Dataset with CNN

Intuition behind MNIST CNN and Tensorspace.js Visualization

CNN Explained: Explorations with CIFAR-10 Dataset

Dropout in CNN: Enhancing Custom Image Classification

Deploying CNN Models: Heroku, AWS, Azure

Deploying CNN Models: GCP, AWS Elastic Beanstalk

Exploring LeNet-5: A Pioneering CNN Architecture

Practical Implementation of LeNet-5: Hands-on Exercises

Unveiling AlexNet: Revolutionizing CNN Architectures

Hands-on with AlexNet: Implementing and Fine-tuning the Model

Delving into VGGNet: Deep CNN with Uniform Architecture

VGG16 Practical: Applying VGGNet for Image Classification Tasks

Introduction to Inception: GoogLeNet's Innovative CNN Architecture

Hands-on with Inception: Implementing InceptionNet for Complex Visual Recognition

Discovering ResNet: Advancements in Deep Residual Learning

ResNet Practical: Building and Fine-tuning Residual Networks for Image Classification

Leveraging Keras Tuner for Hyperparameter Optimization

Building a Simple Model for Image Classification

Fine-tuning Image Classification Models with Keras Tuner

Understanding Data Augmentation: Concepts and Techniques

Advantages of Data Augmentation in Machine Learning

Exploring RICAP, Random Erasing, and Cutout: Advanced Data Augmentation Techniques

Augmentor: An in-depth Exploration of Data Augmentation Library

Roboflow: Harnessing Data Augmentation with Roboflow Platform

Object Detection: Definition, Techniques, and Applications

Object Detection Competitions: Assessing Performance and Advancements

Bounding Boxes in Object Detection: Definition and Significance

Bounding Box Regression: Improving Accuracy in Object Localization

Intersection over Union (IoU): Evaluating Object Detection Performance

Precision and Recall in Object Detection: Key Metrics for Evaluation

Average Precision in Object Detection: Assessing Detection Quality

Key Architectures for Object Detection

R-CNN: Region-based Convolutional Neural Network for Object Detection

The Architecture of R-CNN: Understanding the Network Structure

Cons of R-CNN: Limitations and Challenges

Fast R-CNN: Accelerating Object Detection with Improved Architecture

Architecture of Fast R-CNN: Enhancements and Workflow

Cons of Fast R-CNN: Addressing Challenges and Drawbacks

Faster R-CNN: Advancements in Object Detection Speed and Accuracy

Faster R-CNN Network Architecture: Streamlining Object Detection Pipeline

YOLO: You Only Look Once - Real-Time Object Detection Architecture

YOLO Architecture: One-Stage Detection and Design Principles

YOLO Limitations: Challenges and Trade-offs

SSD: Single Shot MultiBox Detector - Efficient Object Detection Architecture

SSD Network: Structure and Features for Accurate Object Detection

Introduction to TensorFlow Object Detection 1.x (TFOD1.x)

Utilizing Google Colab and Google Drive for Object Detection

Installing Libraries for Object Detection in Colab

Setting up TFOD1.x in Colab for Object Detection

Exploring the Model Zoo for Pretrained Object Detection Models

Performing Object Detection Inference in Colab

Object Detection Inference on Local Machine

Understanding and Configuring Important Files for Object Detection

Testing Object Detection with Webcam Integration

Custom Model Training in TensorFlow Object Detection 1.x (TFOD1.x)

Creating and Organizing Our Custom Dataset for Cards Detection

Annotation or Labeling of Data for Training

Selecting a Pretrained Model from the Model Zoo for Transfer Learning

Setting up Files for Training the Custom Model

Initiating Training in Google Colab

Exporting the Trained Model as a Frozen Inference Graph

Performing Object Detection Inference with our Trained Model in Google Colab

Training the Custom Model on a Local Machine

Object Detection Inference with our Trained Model in a Local Environment

Setting up a PyCharm Project and Environment for Web App Development

Web App Workflow: Understanding the Development Process

Code Understanding: Exploring the Implementation Details

Performing Predictions with Postman: Testing the Web App API

Debugging our Application: Identifying and Resolving Issues

Introduction to TensorFlow Object Detection 2.x (TFOD2.x)

Utilizing the Default Colab Notebook for Object Detection

Setting up Google Colab and Google Drive for TFOD2.x

Exploring the TFOD2.x Model Garden for Pretrained Models

Performing Object Detection Inference using a Pretrained Model in Colab

Object Detection Inference on Local Machine using a Pretrained Model

Custom Model Training in TensorFlow Object Detection 2.x (TFOD2.x)

Creating and Preparing Our Custom Dataset for Chess Piece Detection

Setting up Files and Directories for Model Training

Initiating Model Training: Getting Started with Training

Managing Training: Stopping or Resuming Training Sessions

Evaluating the Trained Model: Assessing Detection Performance

Converting Checkpoint (CKPT) to Saved Model Format

Object Detection Inference with the Custom Trained Model in Google Colab

Object Detection Inference with the Custom Trained Model on a Local PC

Creating a Pycharm project & Environment Setup

Application Workflow

Code understanding

Testing our App with PoSTmaN

Debugging our Application

Introduction to Detectron2

Detectron2 Colab Setup

Visiting Detectron2 Model Zoo

Detectron2 Pretrained Model Inferencing

Detectron2 Custom Training

Exploring the Dataset

Registering Dataset for Training

Let's start Training

Inferencing using the Custom Trained Model in Colab

Evaluating the Model

Creating a Pycharm project & Environment Setup

Application Workflow

Code understanding

Testing our App with Postman

Debugging our Application

Introduction to YoloV5

YoloV5 Colab Setup

Inferencing using Pre Trained Model

Custom Training with YoloV5

Exploring the Dataset

Doing Annotations or labeling data

Setting up Google Colab & Drive

Let's start Training

Inferencing using the Custom Trained Model in Colab

Creating a Pycharm project & Environment Setup

Application Workflow

Code understanding

Testing our App with Postman

Debugging our Application

From Bounding Box to Polygon Masks: Advancements in Image Segmentation Techniques

Image Segmentation: Definition and Applications

Exploring Different Types of Image Segmentation Techniques

MASKRCNN: An Overview of Mask Region-based Convolutional Neural Networks

MASK RCNN Architecture: Key Components and Workflow

Image Segmentation using TFOD1.x: A Hands-on Approach

Local Setup for Mask RCNN Practicals with TFOD

Dataset Exploration for Mask RCNN

Data Annotation for Image Segmentation Training

Model Selection for Mask RCNN in TFOD

Files Setup for Training the Mask RCNN Model

Training the Mask RCNN Model

Exporting the Trained Model as a Frozen Inference Graph

Model Prediction with the Trained Mask RCNN Model

Introduction to Detectron2: A Powerful Framework for Mask RCNN

Utilizing the Detectron2 Colab Notebook for Mask RCNN

Exploring the Model Zoo in Detectron2 for Pretrained Models

Setting up Detectron2 in Colab: Installation and Configuration

Custom Training with Detectron2: Training a Mask RCNN Model

Dataset Exploration in Detectron2: Understanding the Data

Data Annotation for Mask RCNN Training in Detectron2

Data Preparation for Mask RCNN Training: Preparing the Dataset

Setup for Training the Mask RCNN Model in Detectron2

Initiating Model Training: Let's Start Training the Mask RCNN Model

Object Detection Inference with the Custom Trained Model in Colab

Evaluating the Model: Assessing the Performance of the Mask RCNN Model

Installation of Object Tracking Project: Setting up the Environment

Project Demo: Showcasing Object Tracking in Action

Understanding the Code: Exploring the Implementation Details

Understanding the Fundamentals of GANs

Advanced Concepts and Techniques in GANs

GAN Implementation: Step-by-Step Guide to Implementing GANs

GAN Architecture: Exploring Different Architectures in GANs

GAN Practicals: Hands-on Implementation of GANs in Real-world Scenarios

GANS

Instructors

Vincent Thomas Varghese
Vincent Thomas Varghese

Experienced Technostragist with three decades in IT, excelling in Sales, Product Management, and Marketing across IT hardware, networking, and software. Proven strategic planner, startup pioneer, and mentor for large teams. Marketing authority in areas such as Business Incubation, Branding, and Sales. Vast industry expertise spans ITES, IT Hardware, IT Training (Software), Distribution, and Retail. Recognized for interpersonal leadership, intuitive decision-making, and a collaborative approach. Complemented by four years as a Data Scientist, enhancing analytical and problem-solving skills with a deep passion for coding and a knack for simplifying complex concepts,

image not found
₹5500.00
  • Modules
    25 Modules
  • Duration
    65 Hours
  • Category
    Data Science

Login to Purchase the Course