Computer Vision - Fall 2023

University of Central Florida via YouTube

Start learning Write review

Explore comprehensive computer vision concepts through this university-level course covering fundamental image processing techniques, neural networks, and advanced deep learning applications. Master linear algebra foundations essential for computer vision, including vectors, matrices, transformations, and singular value decomposition. Learn image filtering operations using various kernels, Gaussian filters, and noise reduction techniques, then advance to edge detection methods including Prewitt, Sobel, Marr-Hildreth, and Canny edge detectors. Understand neural network fundamentals from basic perceptrons to complex architectures, with detailed coverage of convolutional neural networks (CNNs) including convolution operations, pooling, padding, and stride concepts. Study prominent CNN architectures like AlexNet, ResNet, GoogleNet Inception, and DenseNet while learning training procedures including gradient descent, backpropagation, regularization, and dropout techniques. Gain hands-on experience with PyTorch through comprehensive tutorials covering tensors, computational graphs, automatic differentiation, and building neural networks from scratch. Explore feature extraction methods including corner detection, Harris corner detection, Histogram of Oriented Gradients (HOG), and Scale-Invariant Feature Transform (SIFT) for robust image analysis. Delve into autoencoders for unsupervised feature learning and dimensionality reduction applications. Master classification techniques from traditional methods like k-nearest neighbors and support vector machines to modern fully convolutional networks with various activation functions and loss functions. Advance to object detection using sliding window approaches, template matching, and CNN-based methods including R-CNN and Fast R-CNN with evaluation metrics like Intersection over Union (IoU) and mean Average Precision (mAP). Study image segmentation techniques from basic thresholding and Otsu's method to advanced clustering algorithms including k-means, SLIC, and mean shift segmentation. Conclude with semantic segmentation using fully convolutional networks, U-Net architectures, and skip connections, plus instance segmentation with Mask R-CNN and RoIAlign techniques for precise object boundary detection.

Syllabus

Lecture 1 - Introduction to Computer Vision
Lecture 1.1 - Introduction to Computer Vision [Basics]
Lecture 1.2 - Introduction to Computer Vision [Digital Image]
Lecture 1.3 - Introduction to Computer Vision [Motivation]
Lecture 1.4 - Introduction to Computer Vision [Object Detection]
Lecture 1.5 - Introduction to Computer Vision [Complex Computer Vision Tasks]
Lecture 2 - Linear Algebra Basics
Lecture 2.1 Linear Algebra Basics [Vectors]
Lecture 2.2 - Linear Algebra Basics [Matrix]
Lecture 2.3 - Linear Algebra Basics [Transformations]
Lecture 2.4 - Linear Algebra Basics [Single Value Decomposition]
Lecture 2.5 - Linear Algebra Basics [Q&A]
Lecture 3 - Image Filtering
Lecture 3.1 - Image Filtering [Digitization]
Lecture 3.2 - Image Filtering [Digitization of 1D, 2D, 3D, and Arc]
Lecture 3.3 Image Filtering [Definition, Sampling, Quantization, Examples]
Lecture 3.4 - Image Filtering [Histogram]
Lecture 3.5 - Image Filtering [Intensity Profiles]
Lecture 3.6 - Image Filtering [Image Noise]
Lecture 3.7 - Image Filtering [Why Image Filtering?]
Lecture 3.8 - Image Filtering [Image Filtering Operations using Kernel]
Lecture 3.9 - Image Filtering [Image Filtering Techniques]
Lecture 3.10 - Image Filtering [Gaussian Filter]
Lecture 3.11 - Image Filtering [Filtering Examples]
Lecture 3.12 - Image Filtering [Box Filter]
Lecture 3.13 - Image Filtering [Sobel Filter, and Image Filtering Properties]
Lecture 3.14 - Image Filtering [Median Filter and Kernel size used in Image Filtering]
Lecture 3.15 - Image Filtering [Introduction to Image Derivatives]
Lecture 4 - Edge Detection
Lecture 4.1- Edge Detection [Introduction to Edge Detection]
Lecture 4.2 - Edge Detection [Types of Edges]
Lecture 4.3 - Edge Detection [Characterizing Edges]
Lecture 4.4 - Edge Detection [Effects of Noise in Edge Detection]
Lecture 4.5 - Edge Detection [Smoothing in Edge Detection]
Lecture 4.6 - Edge Detection [Evaluate Edge Detection]
Lecture 4.7 - Edge Detection [Design Criteria of Edge Detection]
Lecture 4.8 - Edge Detection [Introduction to Prewitt and Sobel Edge Detection]
Lecture 4.9 - Edge Detection [Prewitt Edge Detector]
Lecture 4.10 - Edge Detection [Sobel Edge Detector]
Lecture 4.11 - Edge Detection [Marr Hildreth Edge Detector]
Lecture 4.12 - Edge Detection [Laplacian of Gaussian Filtering in Edge Detection]
Lecture 4.13 - Edge Detection [Canny Edge Detection]
Lecture 4.14 Edge Detection [Non Maximum Suppression]
Lecture 4.15 - Edge Detection [Hysteresis Thresholding]
Lecture 5 - Introduction to Neural Networks
Lecture 5.1 - Introduction to Neural Networks [Object Classification]
Lecture 5.2 - Introduction to Neural Networks [Pixel Representation of images]
Lecture 5.3 - Introduction to Neural Networks [Feature Representations]
Lecture 5.4 - Introduction to Neural Networks [Training and Testing Phases in Neural Networks]
Lecture 5.5 - Introduction to Neural Networks [Features used in Neural Networks]
Lecture 5.6 - Introduction to Neural Networks [The Machine Learning Framework]
Lecture 5.7 - Introduction to Neural Networks [Neurons in the Brain 1]
Lecture 5.8 - Introduction to Neural Networks [Neural Activation Function]
Lecture 5.9 - Introduction to Neural Networks [Classifying an Image in Neural Networks]
Lecture 5.10 - Introduction to Neural Networks [Problems with all linear functions]
Lecture 5.11 - Introduction to Neural Networks [Introduction to Non linearities]
Lecture 5.12 - Introduction to Neural Networks [Perceptron]
Lecture 5.13 - Introduction to Neural Networks [Why CNN]
Lecture 6 - Introduction to Convolutional Neural Networks
Lecture 6.1 - Introduction to Convolutional Neural Networks [Motivation for CNN]
Lecture 6.2 - Introduction to Convolutional Neural Networks [Overview of CNN]
Lecture 6.3 - Introduction to Convolutional Neural Networks [Basics]
Lecture 6.4 - Introduction to Convolutional Neural Networks [Fundamental Operation]
Lecture 6.5 - Introduction to Convolutional Neural Networks [Convolution Recap]
Lecture 6.6 - Introduction to Convolutional Neural Networks [Parameters]
Lecture 6.7 - Introduction to Convolutional Neural Networks Convolution [Operation]
Lecture 6.8 - Intro to Convolutional Neural Networks [Understanding Convolution in Sobel edge d]
Lecture 6.9 - Introduction to Convolutional Neural Networks [Convolution Intuition]
Lecture 6.10 - Introduction to Convolutional Neural Networks [2D Convolution]
Lecture 6.11 - Introduction to Convolutional Neural Networks [Stride in 2D Convolution]
Lecture 6.12 - Introduction to Convolutional Neural Networks [Padding]
Lecture 6.13 - Introduction to Convolutional Neural Networks [Pooling]
Lecture 6.14 - Introduction to Convolutional Neural Networks [AlexNet Case Study]
Lecture 6.15 - Introduction to Convolutional Neural Networks [AlexNet Architecture]
Lecture 6.16 - Introduction to Convolutional Neural Networks [Visualizing CNN]
Lecture 7 - Training Neural Networks
Lecture 7.1 - Training Neural Networks [Basics Recap]
Lecture 7.2 Training Neural Networks [Loss Function]
Lecture 7.3 - Training Neural Networks [Gradient Descent in Neural Networks]
Lecture 7.4 - Training Neural Networks [Train CNN with Gradient Descent]
Lecture 7.5 - Training Neural Networks [Differentiability in Neural Networks]
Lecture 7.6 - Training Neural Networks [Backpropagation Chain Rule]
Lecture 7.7 - Training Neural Networks [Stochastic Gradient Descent]
Lecture 7.8 - Training Neural Networks [Stochastic Gradient Descent - Part 2]
Lecture 7.9 - Training Neural Networks [Gradient Descent Oscillations]
Lecture 7.10 - Training Neural Networks [Data Fitting Problem]
Lecture 7.11 - Training Neural Networks [Lowering Learning Rate]
Lecture 7.12 - Training Neural Networks [Regularization]
Lecture 7.13 - Training Neural Networks [Regularization Ensemble]
Lecture 7.14 - Training Neural Networks Dropout
Lecture 7.15 - Training Neural Networks [Training Steps]
Lecture 7.16 - Training Neural Networks [AlexNet Training]
Lecture 7.17 - Training Neural Networks [Residual Network]
Lecture 7.18 - Training Neural Networks [GoogleNet Inception]
Lecture 7.19 - Training Neural Networks [DenseNet]
Lecture 8 - PyTorch Tutorial
Lecture 8.1 - PyTorch Tutorial [PyTorch Tensor]
Lecture 8.2 - PyTorch Tutorial [PyTorch Operations]
Lecture 8.3 - PyTorch Tutorial [Torch Tensor vs Numpy Array]
Lecture 8.4 - PyTorch Tutorial [Matrix Multiplication in PyTorch]
Lecture 8.5 - PyTorch Tutorial [Computational Graphs and Automatic Gradient Computation]
Lecture 8.6 - PyTorch Tutorial [Training Procedure]
Lecture 8.7 - PyTorch Tutorial [Building Neural Network]
Lecture 8.8 - PyTorch Tutorial [Defining a Network class]
Lecture 8.9 - PyTorch Tutorial [Defining a Network class part 2]
Lecture 8.10 - PyTorch Tutorial [Defining a CNN Network - Example]
Lecture 8.11 - PyTorch Tutorial [Iterate over a Dataset of Inputs]
Lecture 8.12 - PyTorch Tutorial [Process input through the Network]
Lecture 8.13 - PyTorch Tutorial [Loss Function]
Lecture 8.14 - PyTorch Tutorial [Gradient Computation]
Lecture 8.15 - PyTorch Tutorial [Update Parameters]
Lecture 8.16 - PyTorch Tutorial [Model Training - Full Procedure]
Lecture 9 - Features
Lecture 9.1 - Features [Introduction to Features]
Lecture 9.2 - Features [Types of Features]
Lecture 9.3 - Features [Uses of Features]
Lecture 9.4 - Features [Finding Features in videos]
Lecture 9.5 - Features [Characteristics of good Features]
Lecture 9.6 - Features [Key Points]
Lecture 9.7 - Features [Properties of Interest Points]
Lecture 9.8 - Features [Corner Detection]
Lecture 9.9 - Features [Corner Detection by Auto correlation]
Lecture 9.10 - Features [Corner Detection strategy]
Lecture 9.11 - Features [Harris Corner Detection]
Lecture 9.12 - Features [Histogram of Gradients Edges]
Lecture 9.13 - Features [HOG - Human Detection]
Lecture 9.14 - Features [Histogram of Oriented Gradients]
Lecture 9.15 - Features [Histogram of Oriented Gradients Example]
Lecture 9.16 - Features [Summary of HOG Computation]
Lecture 9.17 - Features [Introduction to SIFT with an Example]
Lecture 9.18 - Features [Overall Procedure of SIFT at a High Level]
Lecture 9.19 - Features [Automatic scale selection]
Lecture 9.20 - Features [Understanding of useful signature function]
Lecture 9.21 - Features [Finding local maxima in given image]
Lecture 9.22 - Features [Optimizing LoG filter]
Lecture 9.23 - Features [SIFT Recap]
Lecture 9.24 - Features [Implementing Difference of Gaussian]
Lecture 9.25 - Features Extracting Feature Orientation
Lecture 9.26 - Features [Normalization of Orientation]
Lecture 9.27 - Features [Understanding formation of local features in window]
Lecture 9.28 - Features [Summary of SIFT Descriptor]
Lecture 10 - Autoencoder
Lecture 10.1 - Autoencoders [Introduction to Autoencoder]
Lecture 10.2 - Autoencoders [Introduction to Autoencoder - CNN]
Lecture 10.3 - Autoencoders [How Autoencoders work - Examples]
Lecture 10.4 - Autoencoders [Feature Learning using Autoencoder]
Lecture 10.5 - Autoencoders [Applications, and Properties of Autoencoder]
Lecture 11 - Classification I
Lecture 11.1 Classification I [Introduction to classification]
Lecture 11.2 - Classification I [Types of classification]
Lecture 11.3 - Classification I [Understanding how to Perform classification for the Given Data]
Lecture 11.4 - Classification I [General Framework, Decision Boundaries for classification]
Lecture 11.5 - Classification I [Nearest Neighbor Classifier]
Lecture 11.6 - Classification I [K-Nearest Neighbor]
Lecture 11.7 - Classification I [Maximum Margin Classifier - SVM]
Lecture 11.8 - Classification I [Understanding Non separable case]
Lecture 11.9 - Classification I [Support Vector Classifier - Separable case]
Lecture 11.10 - Classification I [Support Vector Classifier - Soft Margin]
Lecture 11.11 - Classification I [Support Vector Classifier Example]
Lecture 11.12 - Classification I [Disadvantages of linear boundary]
Lecture 11.13 - Classification I [Advantages of non linear boundary]
Lecture 11.14 - Classification I [Non linear SVM]
Lecture 11.15 - Classification I [Multi class SVM]
Lecture 11.16 - Classification I [Pros and Cons of SVM]
Lecture 12 - Classification II
Lecture 12.1 - Classification II [Classification I Recap]
Lecture 12.2 - Classification II [Fully Convolutional Network]
Lecture 12.3 - Classification II [Converting Fully Convolution to Convolution]
Lecture 12.4 - Classification II [Introducing Non linearities]
Lecture 12.5 - Classification II [Activation Functions]
Lecture 12.6 - Classification II [Softmax Classification]
Lecture 12.7 - Classification II [Loss Function, Visualizing Convolution]
Lecture 13 - Object Detection
Lecture 13.1 - Object Detection {Introduction to Object Detection]
Lecture 13.2 - Object Detection [Introduction to Sliding Window, Template Matching]
Lecture 13.3 - Object Detection [Example of Sliding Window Approach]
Lecture 13.4 - Object Detection [Gaussian pre-filtering]
Lecture 13.5 - Object Detection [Gaussian Pyramid Construction and Challenges]
Lecture 13.6 - Object Detection [Extract Features, Classify Features, Post Processing]
Lecture 13.7 - Object Detection [Non maximum suppression]
Lecture 13.8 - Object Detection [Evaluation using IOU]
Lecture 13.9 - Object Detection [Precision and Recall]
Lecture 13.10 - Object Detection [Mean Average Precision]
Lecture 13.11 - Object Detection [CNN Based Approach - A Simple Solution]
Lecture 13.12 - Object Detection [CNN based Approach RCNN]
Lecture 13.13 - Object Detection [CNN based Approach Fast R-CNN]
Lecture 13.14 - Object Detection [Per class or class agnostic regression]
Lecture 14 - Image Segmentation
Lecture 14.2 - Image Segmentation [Image Binarization]
Lecture 14.3 - Image Segmentation [Thresholding Examples]
Lecture 14.4 - Image Segmentation [Otsu Example]
Lecture 14.5 - Image Segmentation [Region Growing Algorithm]
Lecture 14.6 - Image Segmentation [Region Splitting and Merging 1]
Lecture 14.7 - Image Segmentation [Introduction to Clustering]
Lecture 14.8 - Image Segmentation [Natural Grouping and Distance Metrics]
Lecture 14.9 - Image Segmentation [K-means clustering algorithm]
Lecture 14.10 - Image Segmentation [K-means Loss Function]
Lecture 14.11 - Image Segmentation [SLIC Algorithm]
Lecture 14.12 - Image Segmentation [Mean Shift Segmentation]
Lecture 14.13 - Image Segmentation [Mean Shift Algorithm]
Lecture 14.14 - Image Segmentation [Mean Shift Clustering]
Lecture 15 - Semantic Segmentation
Lecture 15.1 - Semantic Segmentation [Introduction to Semantic Segmentation]
Lecture 15.2 - Semantic Segmentation [Segmentation tasks]
Lecture 15.3 - Semantic Segmentation [Fully Convolutional Networks]
Lecture 15.4 - Semantic Segmentation [Skip Connections in Fully Convolutional Networks]
Lecture 15.5 - Semantic Segmentation [Upsampling]
Lecture 15.6 - Semantic Segmentation [Deconvolution network for Semantic Segmentation]
Lecture 15.7 - Semantic Segmentation [U-Net Sampling]
Lecture 16 - Instance Segmentation
Lecture 16.1 - Instance Segmentation [Introduction to Instance Segmentation]
Lecture 16.2 - Instance Segmentation [Mask - RCNN]
Lecture 16.3 - Instance Segmentation [RoIAlign vs RoIPool]
Lecture 16.4 - Instance Segmentation [RoI-Align vs RoI-Pooling Example]
Lecture 16.5 - Instance Segmentation [Mask - RCNN Architecture]