This is a showcase of my work in the course Automatic Image Analysis at the Technical University Berlin for my Master's in Computer Science. I applied a few different neural network structures to the FashionMnist dataset for classification. Specifically, I have designed a ...
- Base Model (a small Convolutional neural net in the classic sense)
- Batch Norm Model (The base model with batch norm in addition)
- Deeper Model (a deeper neural net with Batch Norm than the base model)
- Residual Model (a neural net with residual connections)