Gain insights into using PyTorch for image classification and detection, delve into model implementation, and explore deployment on edge devices with ONNX and OpenVINO.

Image Classification and Object Detection using CNNs-01.png

pytorch.tar.gz

Basics_Of_CNN-ex1

Basics_Of_CNN-ex2

Basics_Of_CNN-vgg16

Basics_Of_CNN-inceptionv1

Basics_Of_CNN-residuals

Basics_Of_CNN-depthwise

Train-data

Train-training

Train-trainingFT

Train-transfer

deploy-onnx

deploy-openvino

deploy-comparison

prepare_dataset_openimages

train_yolov7

inference_yolov7

onnx_yolov7

Image classification and object detection have gained widespread use in recent years. Content categorization and monitoring, disease diagnosis from medical images, identifying terrain in satellite images, and detecting road elements for self-driving cars are classification problems at their core. PyTorch is a popular framework for these tasks—offering a useful mix of user-friendliness, deep learning functionalities, customization, and optimization.

In this course, you will cover the fundamentals of classification and object detection models and apply them to actual datasets using PyTorch. You’ll learn popular architectures and how to implement and fine-tune them for better results. Finally, you’ll learn to convert models to ONNX and OpenVINO to deploy in edge devices.

By the end of this course, you will have acquired the necessary skills to be able to use PyTorch for image identification and object detection in real-world applications

Using PyTorch for Image Classification and Object Detection

We will learn how to write code to prepare our dataset and train our model to obtain the final weights and use it during inference in the next step. The flow is shown below:



# Data preparation

Data preparation is required to obtain the appropriate format of the data and send it through the model during training. For this purpose, we will learn how to:

* Split the dataset into train, test, and validation subsets.
* Apply data augmentation to the default dataset.
* Create a custom dataset module.


# Training

Building a neural network architecture is just the beginning of obtaining a model we can use in our projects, predicting the necessary information. Training the created network is the most important and challenging part, and we need to have good knowledge about various points. For this purpose, we will learn how to:

* Create our neural network.
* Send our data through the model.
* Calculate the loss after each epoch. 
* Apply backpropagation using gradient descent and update the weights.
* Plot training and validation accuracy to track model performance.
* Save the trained model.

# Testing

A good model should perform robustly on data never seen during training. To measure the performance of our trained model on the new data, we will learn how to create a test pipeline.

# Fine-tuning

In many projects, whether we have a complex or simple dataset, it is expected that the first training doesn't give the best result that we can obtain. This is because every dataset and classification problem we work on is unique. Therefore, we need to be able to understand what the weak points of our network are, and how we can improve it. For this purpose, we will learn how to:

* Track the gradients.
* Tackle overfitting-underfitting problems using various regularizations like batch normalization, dropout, L1 & L2 regularization, etc.
* Transfer learning.

# Inferencing

Inference refers to when we are done with training and improving our model and are ready to use it.

Get an overview of what we’ll cover in this chapter.


Introduction of Using PyTorch for Image Classification

Introduction to PyTorch for Image Classification

Before We Start

Basics of Convolutional Neural Networks

Cats vs Dogs Classification with Convolutional Neural Networks

Popular Neural Network Architectures for Image Classification

Using PyTorch for Image Classification

Model Deployment

Using a PyTorch Model in JavaScript with ONNX

Basics of Object Detection

Two-Stage Object Detection Architectures

One-Stage Object Detection Architectures

YOLOv7 Model Train and Inference on Edge

Conclusion

Appendix

Building a System for Safety Helmet Detection Based on YOLOv5

Introduction to PyTorch for Image Classification

Data preparation