This is one of my homework assignments I did in the courses I have taken.
Although this project is not built from scratch, I did the entirety of the significant implementations that makes this project function properly.
Here is what I did
- Downloaded sign language MNIST data set
- Parsed data from CSV file into NumPy arrays
- Created ImageDataGenerator objects to create test and validation data my neural network could actually train on. Implemented data augmentation to increase the amount of testing data.
- Built a convolutional neural network and trained it eventually reaching a validation accuracy of 96%
Skills gained:
- Parsing data in CSV files
- Data Augmentation
- Building highly accurate CNNs