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

Concatenation requires the channel sizes to be equal; therefore, this attempt would give us an error.


Considering we have two input tensors with channel sizes 10 and 20, what would be the output feature map channel after concatenation?

Considering we have two input tensors with channel sizes 10 and 20, what would be the output feature map channel after concatenation?

Data augmentation techniques specific to object detection, like Mosaic, are in the Bag of Specials group.


When we have different sizes of implementations for the same object detector, it's always better to use the biggest one.

If we see an operation visualized as $\bigoplus$, it's refers to concatenation.

Regardless of using different types of backbones, necks, or heads, if a model completes region proposal and classification steps in one stage, it’s a one-stage detector.


Which statement is true?

The branch number is a dimension for architectures that aggregate feature paths in a feature pyramidal way. It is called a stage and is considered in compound scaling in YOLOv7.


If we sum two input tensors from channel sizes 5 and 10, we get the output feature map with 15 channels.

Model reparametrization can be applied by training the same architecture multiple times with different datasets and taking the average of the weights from these trainings. 

An object detector should have anchor boxes to produce region proposals.

**(Select all that apply.)** Which expression is _true_?

(Select all that apply.) Which expression is true?

To use residual-like blocks in training and inference time


To use residual-like blocks in inference time and default VGG blocks in training

To use residual-like blocks in training time and default VGG blocks in inference

What is the main working logic of RepVGG?

What is the main working logic of RepVGG?

All the members use anchor boxes to create ROIs.

YOLOv7 brings the compound scaling method to concatenation-based architectures, scaling the number of feature pyramids in a branch along with resolution, width, and height.

Most of the YOLO architectures are available in different sizes, including nano, tiny, small, medium, and large.


**(Select all that apply.)** Which statement is _true_ about the YOLO object detection family?

(Select all that apply.) Which statement is true about the YOLO object detection family?

Test your knowledge and understanding of one-stage object detectors.

Before We Start

Basics of Convolutional Neural Networks

Popular Neural Network Architectures for Image Classification

Using PyTorch for Image Classification

Model Deployment

Basics of Object Detection

Two-Stage Object Detection Architectures

One-Stage Object Detection Architectures

YOLOv7 Model Train and Inference on Edge

Conclusion

Appendix

Quiz on One-Stage Object Detectors