A1.
- Implementing a neural network from scratch using Numpy toolbox.
A2.
- Saliency map prediction of images by Deep Convnet model.
- Reconstraction of images by thier feature maps from different layers of AlexNet.
A3.
- Brain tumor tissue seperation in MRI pictures by image segmentation with U-Net model.
- Part-Of-Speech (POS) tagging with RNNs.
A4.
- Using DistilHuBert transformer model for speech-to-text and keyword spotting tasks
- Using DistilBert and XLM transformer models for question-answering task
The main framework used for almost every assignment is Pytorch.