Graduate project for CS200A(DS100) at UC Berkeley. Training a classifier for 20 classes using 1500 images.
-
Data Cleaning
-
Reading and categorizing the image files
-
Using Scikit-Image
-
-
Feature Engineering
-
Exploratory data analysis of image classification features
-
Produce insightful graphs
-
-
Model Training
-
Develop models to classify images
-
Experimenting with different machine-learning methods
-
-
TensorFlow
- Explore the future of computer vision