This repo contains exercises and solutions for the Computer Vision Pipelines lesson in Udacity's Data Science Nanodegree.
The repo contains the following demos and exercises:
- Preprocessing and transforming image data in OpenCV
- Extracting image features like gradients, edges and corners in OpenCV
- Classifying images of fashion items with support vector machines in scikit-learn
- Wrapping OpenCV in a custom scikit-learn transformer
- Optimizing the full computer vision pipeline including SVM parameters and OpenCV image features
- Implementing a convolutional neural network in PyTorch and using it within scikit-learn pipelines
See Pipfile
for an overview of required python packages.
Antje Muntzinger
(c) 2024, Udacity Inc.