/ComputerVision

The repository is dedicated towards the implementation of Computer Vision.

Primary LanguageJupyter NotebookMIT LicenseMIT

COMPUTER VISION

The repository contains a list of projects and notebooks which I have worked on while reading Computer Vision, Deep Learning and OpenCV from PyImageSearch.

đź“šNOTEBOOKS:

1. OPEN CV

  • The OpenCV notebook contains the basics of OpenCV such as Loading an Image, Resizing Images, Rotating Image, Edge Detection, Thresholding, Drawing and Masking, Contour and Shape Detection. The OCV Project I notebook contains the implementation of Rotating Images correctly without cut off. The OCV Project II notebook contains the implementation of Color Detection and Histogram Matching on images.

2. CONVOLUTIONAL NEURAL NETWORKS

  • The Convolutions notebook contains all the dependencies required to understand the implementation of Image Convolutions and Kernels. The Convolutional Layers notebook contains all the dependencies required to understand Keras Conv2D Class and Convolutional Layers.

3. IMAGE CLASSIFICATION

  • The Simplepreprocessor notebook contains implementation of simple image preprocessor, loading an image dataset into memory, and K-Nearest Neighbor Classifier. K-Nearest Neighbor Classifier doesn’t actually learn anything, but it directly relies on the distance between feature vectors. The GradientDescent notebook contains implementation of Gradient Descent Algorithms. The StochasticGradientDescent notebook contains implementation of Stochastic Gradient Descent, Image Classification and Regularization.

4. NEURAL NETWORKS

  • The Neural Networks notebook contains the implementation of Perceptron algorithm, backpropagation algorithm, and neural networks from scratch.

5. LENET ARCHITECTURE

  • The LeNet Architecture notebook contains the implementation of LeNet Architecture. LeNet is a seminal work in the deep learning literature which demonstrates how neural networks could be trained to recognize objects in images without feture extraction.

6. VGGNET ARCHITECTURE

  • The Mini VGGNet notebook contains the implementation of VGGNet Architecture. It makes the use of only 3 X 3 filters regardless of network depth.

7. PRETRAINED CNN

  • The Pretrained CNNs notebook contains the reviews of convolutional neural networks and implementation of VGG16 and Xception networks for classification of images.

8. OBJECT DETECTION

  • The Object Detection notebook contains implementation of object detection with PyTorch and pretrained networks. It also contains brief description about object detection and image preprocessing.